From 27f39c73e63833b4c081a0d681d88b4184a0491d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 19 May 2010 22:07:23 +0000 Subject: net: Use __this_cpu_inc() in fast path This patch saves 224 bytes of text on my machine. __this_cpu_inc() generates a single instruction, using no scratch registers : 65 ff 04 25 a8 30 01 00 incl %gs:0x130a8 instead of : 48 c7 c2 80 30 01 00 mov $0x13080,%rdx 65 48 8b 04 25 88 ea 00 00 mov %gs:0xea88,%rax 83 44 10 28 01 addl $0x1,0x28(%rax,%rdx,1) Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 3 +-- net/ipv4/route.c | 3 +-- 2 files changed, 2 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/core/dev.c b/net/core/dev.c index 1845b08..7d76b05 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2822,8 +2822,7 @@ static int __netif_receive_skb(struct sk_buff *skb) skb->dev = master; } - __get_cpu_var(softnet_data).processed++; - + __this_cpu_inc(softnet_data.processed); skb_reset_network_header(skb); skb_reset_transport_header(skb); skb->mac_len = skb->network_header - skb->mac_header; diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 560acc6..8495bce 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -253,8 +253,7 @@ static unsigned rt_hash_mask __read_mostly; static unsigned int rt_hash_log __read_mostly; static DEFINE_PER_CPU(struct rt_cache_stat, rt_cache_stat); -#define RT_CACHE_STAT_INC(field) \ - (__raw_get_cpu_var(rt_cache_stat).field++) +#define RT_CACHE_STAT_INC(field) __this_cpu_inc(rt_cache_stat.field) static inline unsigned int rt_hash(__be32 daddr, __be32 saddr, int idx, int genid) -- cgit v1.1 From 15e83ed78864d0625e87a85f09b297c0919a4797 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 19 May 2010 23:16:03 +0000 Subject: net: remove zap_completion_queue netpoll does an interesting work in zap_completion_queue(), but this was before we did skb orphaning before delivering packets to device. It now makes sense to add a test in dev_kfree_skb_irq() to not queue a skb if already orphaned, and to remove netpoll zap_completion_queue() as a bonus. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 4 +++- net/core/netpoll.c | 31 ------------------------------- 2 files changed, 3 insertions(+), 32 deletions(-) (limited to 'net') diff --git a/net/core/dev.c b/net/core/dev.c index 7d76b05..983a3c1 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1577,7 +1577,9 @@ EXPORT_SYMBOL(__netif_schedule); void dev_kfree_skb_irq(struct sk_buff *skb) { - if (atomic_dec_and_test(&skb->users)) { + if (!skb->destructor) + dev_kfree_skb(skb); + else if (atomic_dec_and_test(&skb->users)) { struct softnet_data *sd; unsigned long flags; diff --git a/net/core/netpoll.c b/net/core/netpoll.c index 94825b1..e034342 100644 --- a/net/core/netpoll.c +++ b/net/core/netpoll.c @@ -49,7 +49,6 @@ static atomic_t trapped; (MAX_UDP_CHUNK + sizeof(struct udphdr) + \ sizeof(struct iphdr) + sizeof(struct ethhdr)) -static void zap_completion_queue(void); static void arp_reply(struct sk_buff *skb); static unsigned int carrier_timeout = 4; @@ -197,7 +196,6 @@ void netpoll_poll_dev(struct net_device *dev) service_arp_queue(dev->npinfo); - zap_completion_queue(); } void netpoll_poll(struct netpoll *np) @@ -221,40 +219,11 @@ static void refill_skbs(void) spin_unlock_irqrestore(&skb_pool.lock, flags); } -static void zap_completion_queue(void) -{ - unsigned long flags; - struct softnet_data *sd = &get_cpu_var(softnet_data); - - if (sd->completion_queue) { - struct sk_buff *clist; - - local_irq_save(flags); - clist = sd->completion_queue; - sd->completion_queue = NULL; - local_irq_restore(flags); - - while (clist != NULL) { - struct sk_buff *skb = clist; - clist = clist->next; - if (skb->destructor) { - atomic_inc(&skb->users); - dev_kfree_skb_any(skb); /* put this one back */ - } else { - __kfree_skb(skb); - } - } - } - - put_cpu_var(softnet_data); -} - static struct sk_buff *find_skb(struct netpoll *np, int len, int reserve) { int count = 0; struct sk_buff *skb; - zap_completion_queue(); refill_skbs(); repeat: -- cgit v1.1 From d7fd1b5747fff3bde92777bcaf705d98ae6f8b6f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 17 May 2010 20:40:51 +0000 Subject: tcp: tcp_md5_hash_skb_data() frag_list handling tcp_md5_hash_skb_data() should handle skb->frag_list, and eventually recurse. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'net') diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 6596b4f..49d0d2b 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -2999,6 +2999,7 @@ int tcp_md5_hash_skb_data(struct tcp_md5sig_pool *hp, const unsigned head_data_len = skb_headlen(skb) > header_len ? skb_headlen(skb) - header_len : 0; const struct skb_shared_info *shi = skb_shinfo(skb); + struct sk_buff *frag_iter; sg_init_table(&sg, 1); @@ -3013,6 +3014,10 @@ int tcp_md5_hash_skb_data(struct tcp_md5sig_pool *hp, return 1; } + skb_walk_frags(skb, frag_iter) + if (tcp_md5_hash_skb_data(hp, frag_iter, 0)) + return 1; + return 0; } -- cgit v1.1 From 16dff91804777efbb0ce18b0a7e54c55e86b7beb Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Fri, 21 May 2010 22:18:34 +0000 Subject: net/can: Use memdup_user Use memdup_user when user data is immediately copied into the allocated region. The semantic patch that makes this change is as follows: (http://coccinelle.lip6.fr/) // @@ expression from,to,size,flag; position p; identifier l1,l2; @@ - to = \(kmalloc@p\|kzalloc@p\)(size,flag); + to = memdup_user(from,size); if ( - to==NULL + IS_ERR(to) || ...) { <+... when != goto l1; - -ENOMEM + PTR_ERR(to) ...+> } - if (copy_from_user(to, from, size) != 0) { - <+... when != goto l2; - -EFAULT - ...+> - } // Signed-off-by: Julia Lawall Signed-off-by: David S. Miller --- net/can/raw.c | 11 +++-------- 1 file changed, 3 insertions(+), 8 deletions(-) (limited to 'net') diff --git a/net/can/raw.c b/net/can/raw.c index da99cf1..ccfe633 100644 --- a/net/can/raw.c +++ b/net/can/raw.c @@ -436,14 +436,9 @@ static int raw_setsockopt(struct socket *sock, int level, int optname, if (count > 1) { /* filter does not fit into dfilter => alloc space */ - filter = kmalloc(optlen, GFP_KERNEL); - if (!filter) - return -ENOMEM; - - if (copy_from_user(filter, optval, optlen)) { - kfree(filter); - return -EFAULT; - } + filter = memdup_user(optval, optlen); + if (IS_ERR(filter)) + return PTR_ERR(filter); } else if (count == 1) { if (copy_from_user(&sfilter, optval, sizeof(sfilter))) return -EFAULT; -- cgit v1.1 From 042604d2a3ee59c84f8293988caf35bac5de9eb3 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Fri, 21 May 2010 22:25:19 +0000 Subject: net/dccp: Use memdup_user Use memdup_user when user data is immediately copied into the allocated region. The semantic patch that makes this change is as follows: (http://coccinelle.lip6.fr/) // @@ expression from,to,size,flag; position p; identifier l1,l2; @@ - to = \(kmalloc@p\|kzalloc@p\)(size,flag); + to = memdup_user(from,size); if ( - to==NULL + IS_ERR(to) || ...) { <+... when != goto l1; - -ENOMEM + PTR_ERR(to) ...+> } - if (copy_from_user(to, from, size) != 0) { - <+... when != goto l2; - -EFAULT - ...+> - } // Signed-off-by: Julia Lawall Acked-by: Gerrit Renker Signed-off-by: David S. Miller --- net/dccp/proto.c | 11 +++-------- 1 file changed, 3 insertions(+), 8 deletions(-) (limited to 'net') diff --git a/net/dccp/proto.c b/net/dccp/proto.c index b03ecf6..f79bcef 100644 --- a/net/dccp/proto.c +++ b/net/dccp/proto.c @@ -473,14 +473,9 @@ static int dccp_setsockopt_ccid(struct sock *sk, int type, if (optlen < 1 || optlen > DCCP_FEAT_MAX_SP_VALS) return -EINVAL; - val = kmalloc(optlen, GFP_KERNEL); - if (val == NULL) - return -ENOMEM; - - if (copy_from_user(val, optval, optlen)) { - kfree(val); - return -EFAULT; - } + val = memdup_user(optval, optlen); + if (IS_ERR(val)) + return PTR_ERR(val); lock_sock(sk); if (type == DCCP_SOCKOPT_TX_CCID || type == DCCP_SOCKOPT_CCID) -- cgit v1.1 From f5d72af9f3dfea67f014f880d00665f96622f57f Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 22 May 2010 10:43:42 +0000 Subject: caif: remove unneeded null check in caif_connect() We already dereferenced uaddr towards the start of the function when we checked that "uaddr->sa_family != AF_CAIF". Both the check here and the earlier check were added in bece7b2398d0: "caif: Rewritten socket implementation". Before that patch, we assumed that we recieved a valid pointer for uaddr, and based on that, I have removed this check. Signed-off-by: Dan Carpenter Acked-by: Sjur Braendeland Signed-off-by: David S. Miller --- net/caif/caif_socket.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'net') diff --git a/net/caif/caif_socket.c b/net/caif/caif_socket.c index 3d0e095..7912493 100644 --- a/net/caif/caif_socket.c +++ b/net/caif/caif_socket.c @@ -874,8 +874,7 @@ static int caif_connect(struct socket *sock, struct sockaddr *uaddr, sk_stream_kill_queues(&cf_sk->sk); err = -EINVAL; - if (addr_len != sizeof(struct sockaddr_caif) || - !uaddr) + if (addr_len != sizeof(struct sockaddr_caif)) goto out; memcpy(&cf_sk->conn_req.sockaddr, uaddr, -- cgit v1.1 From 9bfca3c6e5965a01b7c6d8e2f2ea30799c9d43bc Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Mon, 24 May 2010 06:29:34 +0000 Subject: caif: cleanup: remove duplicate checks "phyinfo" can never be null here because we assigned it an address, so I removed both the assert and the second check inside the if statement. I removed the "phyinfo->phy_layer != NULL" check as well because that was asserted earlier. Walter Harms suggested I move the "phyinfo->phy_ref_count++;" outside the if condition for readability, so I have done that. Signed-off-by: Dan Carpenter Acked-by: Sjur Braendeland Signed-off-by: David S. Miller --- net/caif/cfcnfg.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) (limited to 'net') diff --git a/net/caif/cfcnfg.c b/net/caif/cfcnfg.c index df43f26..7c81974 100644 --- a/net/caif/cfcnfg.c +++ b/net/caif/cfcnfg.c @@ -308,19 +308,15 @@ cfcnfg_linkup_rsp(struct cflayer *layer, u8 channel_id, enum cfctrl_srv serv, caif_assert(cnfg != NULL); caif_assert(phyid != 0); phyinfo = &cnfg->phy_layers[phyid]; - caif_assert(phyinfo != NULL); caif_assert(phyinfo->id == phyid); caif_assert(phyinfo->phy_layer != NULL); caif_assert(phyinfo->phy_layer->id == phyid); - if (phyinfo != NULL && - phyinfo->phy_ref_count++ == 0 && - phyinfo->phy_layer != NULL && + phyinfo->phy_ref_count++; + if (phyinfo->phy_ref_count == 1 && phyinfo->phy_layer->modemcmd != NULL) { - caif_assert(phyinfo->phy_layer->id == phyid); phyinfo->phy_layer->modemcmd(phyinfo->phy_layer, _CAIF_MODEMCMD_PHYIF_USEFULL); - } adapt_layer->id = channel_id; -- cgit v1.1 From 06c4648d46d1b757d6b9591a86810be79818b60c Mon Sep 17 00:00:00 2001 From: Ian Campbell Date: Wed, 26 May 2010 00:09:42 +0000 Subject: arp_notify: allow drivers to explicitly request a notification event. Currently such notifications are only generated when the device comes up or the address changes. However one use case for these notifications is to enable faster network recovery after a virtual machine migration (by causing switches to relearn their MAC tables). A migration appears to the network stack as a temporary loss of carrier and therefore does not trigger either of the current conditions. Rather than adding carrier up as a trigger (which can cause issues when interfaces a flapping) simply add an interface which the driver can use to explicitly trigger the notification. Signed-off-by: Ian Campbell Cc: Stephen Hemminger Cc: Jeremy Fitzhardinge Cc: David S. Miller Cc: netdev@vger.kernel.org Cc: stable@kernel.org Signed-off-by: David S. Miller --- net/ipv4/devinet.c | 1 + net/sched/sch_generic.c | 18 ++++++++++++++++++ 2 files changed, 19 insertions(+) (limited to 'net') diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c index 382bc76..da14c49 100644 --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -1081,6 +1081,7 @@ static int inetdev_event(struct notifier_block *this, unsigned long event, } ip_mc_up(in_dev); /* fall through */ + case NETDEV_NOTIFY_PEERS: case NETDEV_CHANGEADDR: /* Send gratuitous ARP to notify of link change */ if (IN_DEV_ARP_NOTIFY(in_dev)) { diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index a63029e..bd1892f 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -327,6 +327,24 @@ void netif_carrier_off(struct net_device *dev) } EXPORT_SYMBOL(netif_carrier_off); +/** + * netif_notify_peers - notify network peers about existence of @dev + * @dev: network device + * + * Generate traffic such that interested network peers are aware of + * @dev, such as by generating a gratuitous ARP. This may be used when + * a device wants to inform the rest of the network about some sort of + * reconfiguration such as a failover event or virtual machine + * migration. + */ +void netif_notify_peers(struct net_device *dev) +{ + rtnl_lock(); + call_netdevice_notifiers(NETDEV_NOTIFY_PEERS, dev); + rtnl_unlock(); +} +EXPORT_SYMBOL(netif_notify_peers); + /* "NOOP" scheduler: the best scheduler, recommended for all interfaces under all circumstances. It is difficult to invent anything faster or cheaper. -- cgit v1.1 From 3ed37a6fa70a3a63dbb257fc640facb3974bba40 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Mon, 31 May 2010 17:23:21 +0000 Subject: net/ipv4/igmp.c: Remove unnecessary kmalloc casts Signed-off-by: Joe Perches Signed-off-by: David S. Miller --- net/ipv4/igmp.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'net') diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index 5fff865..250cb5e 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -1646,8 +1646,7 @@ static int sf_setstate(struct ip_mc_list *pmc) if (dpsf->sf_inaddr == psf->sf_inaddr) break; if (!dpsf) { - dpsf = (struct ip_sf_list *) - kmalloc(sizeof(*dpsf), GFP_ATOMIC); + dpsf = kmalloc(sizeof(*dpsf), GFP_ATOMIC); if (!dpsf) continue; *dpsf = *psf; -- cgit v1.1 From 5d55354f147bcf82b7a330bf9617b82a692b7d49 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Mon, 31 May 2010 17:23:22 +0000 Subject: net/ipv6/mcast.c: Remove unnecessary kmalloc casts Signed-off-by: Joe Perches Signed-off-by: David S. Miller --- net/ipv6/mcast.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'net') diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c index 59f1881..7b5fb43 100644 --- a/net/ipv6/mcast.c +++ b/net/ipv6/mcast.c @@ -1995,8 +1995,7 @@ static int sf_setstate(struct ifmcaddr6 *pmc) &psf->sf_addr)) break; if (!dpsf) { - dpsf = (struct ip6_sf_list *) - kmalloc(sizeof(*dpsf), GFP_ATOMIC); + dpsf = kmalloc(sizeof(*dpsf), GFP_ATOMIC); if (!dpsf) continue; *dpsf = *psf; -- cgit v1.1 From aac4dddc358acfd9d98b20024a42c34dfab31c39 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 1 Jun 2010 00:26:58 -0700 Subject: ipv6: get rid of ipip6_prl_lock As noticed by Julia Lawall, ipip6_tunnel_add_prl() incorrectly calls kzallloc(..., GFP_KERNEL) while a spinlock is held. She provided a patch to use GFP_ATOMIC instead. One possibility would be to convert this spinlock to a mutex, or preallocate the thing before taking the lock. After RCU conversion, it appears we dont need this lock, since caller already holds RTNL Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv6/sit.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) (limited to 'net') diff --git a/net/ipv6/sit.c b/net/ipv6/sit.c index e51e650..702c532 100644 --- a/net/ipv6/sit.c +++ b/net/ipv6/sit.c @@ -249,8 +249,6 @@ failed: return NULL; } -static DEFINE_SPINLOCK(ipip6_prl_lock); - #define for_each_prl_rcu(start) \ for (prl = rcu_dereference(start); \ prl; \ @@ -340,7 +338,7 @@ ipip6_tunnel_add_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a, int chg) if (a->addr == htonl(INADDR_ANY)) return -EINVAL; - spin_lock(&ipip6_prl_lock); + ASSERT_RTNL(); for (p = t->prl; p; p = p->next) { if (p->addr == a->addr) { @@ -370,7 +368,6 @@ ipip6_tunnel_add_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a, int chg) t->prl_count++; rcu_assign_pointer(t->prl, p); out: - spin_unlock(&ipip6_prl_lock); return err; } @@ -397,7 +394,7 @@ ipip6_tunnel_del_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a) struct ip_tunnel_prl_entry *x, **p; int err = 0; - spin_lock(&ipip6_prl_lock); + ASSERT_RTNL(); if (a && a->addr != htonl(INADDR_ANY)) { for (p = &t->prl; *p; p = &(*p)->next) { @@ -419,7 +416,6 @@ ipip6_tunnel_del_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a) } } out: - spin_unlock(&ipip6_prl_lock); return err; } -- cgit v1.1 From a8b563894d6fee9b90b7d6ed76f8ec28ad45bcbe Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 1 Jun 2010 11:48:31 +0200 Subject: netfilter: br_netfilter: use skb_set_noref() Avoid dirtying bridge_parent_rtable refcount, using new dst noref infrastructure. Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/bridge/br_netfilter.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/bridge/br_netfilter.c b/net/bridge/br_netfilter.c index 4442099..cbea5af 100644 --- a/net/bridge/br_netfilter.c +++ b/net/bridge/br_netfilter.c @@ -244,8 +244,7 @@ static int br_nf_pre_routing_finish_ipv6(struct sk_buff *skb) kfree_skb(skb); return 0; } - dst_hold(&rt->u.dst); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set_noref(skb, &rt->u.dst); skb->dev = nf_bridge->physindev; nf_bridge_update_protocol(skb); @@ -396,8 +395,7 @@ bridged_dnat: kfree_skb(skb); return 0; } - dst_hold(&rt->u.dst); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set_noref(skb, &rt->u.dst); } skb->dev = nf_bridge->physindev; -- cgit v1.1 From fabf3a85ab88063c10f367cccba7b3a1e59df996 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 1 Jun 2010 12:00:41 +0200 Subject: netfilter: xt_statistic: remove nth_lock spinlock Use atomic_cmpxchg() to avoid dirtying a shared location. xt_statistic_priv smp aligned to avoid sharing same cache line with other stuff. Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/netfilter/xt_statistic.c | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-) (limited to 'net') diff --git a/net/netfilter/xt_statistic.c b/net/netfilter/xt_statistic.c index 96e62b8..42ecb71 100644 --- a/net/netfilter/xt_statistic.c +++ b/net/netfilter/xt_statistic.c @@ -18,8 +18,8 @@ #include struct xt_statistic_priv { - uint32_t count; -}; + atomic_t count; +} ____cacheline_aligned_in_smp; MODULE_LICENSE("GPL"); MODULE_AUTHOR("Patrick McHardy "); @@ -27,13 +27,12 @@ MODULE_DESCRIPTION("Xtables: statistics-based matching (\"Nth\", random)"); MODULE_ALIAS("ipt_statistic"); MODULE_ALIAS("ip6t_statistic"); -static DEFINE_SPINLOCK(nth_lock); - static bool statistic_mt(const struct sk_buff *skb, struct xt_action_param *par) { const struct xt_statistic_info *info = par->matchinfo; bool ret = info->flags & XT_STATISTIC_INVERT; + int nval, oval; switch (info->mode) { case XT_STATISTIC_MODE_RANDOM: @@ -41,12 +40,12 @@ statistic_mt(const struct sk_buff *skb, struct xt_action_param *par) ret = !ret; break; case XT_STATISTIC_MODE_NTH: - spin_lock_bh(&nth_lock); - if (info->master->count++ == info->u.nth.every) { - info->master->count = 0; + do { + oval = atomic_read(&info->master->count); + nval = (oval == info->u.nth.every) ? 0 : oval + 1; + } while (atomic_cmpxchg(&info->master->count, oval, nval) != oval); + if (nval == 0) ret = !ret; - } - spin_unlock_bh(&nth_lock); break; } @@ -64,7 +63,7 @@ static int statistic_mt_check(const struct xt_mtchk_param *par) info->master = kzalloc(sizeof(*info->master), GFP_KERNEL); if (info->master == NULL) return -ENOMEM; - info->master->count = info->u.nth.count; + atomic_set(&info->master->count, info->u.nth.count); return 0; } -- cgit v1.1 From bc135b23d01acf7ee926aaf75b0020c86d3869f9 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 2 Jun 2010 03:23:51 -0700 Subject: net: Define accessors to manipulate QDISC_STATE_RUNNING Define three helpers to manipulate QDISC_STATE_RUNNIG flag, that a second patch will move on another location. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 4 ++-- net/sched/sch_generic.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/core/dev.c b/net/core/dev.c index 983a3c1..2733226 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2047,7 +2047,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, kfree_skb(skb); rc = NET_XMIT_DROP; } else if ((q->flags & TCQ_F_CAN_BYPASS) && !qdisc_qlen(q) && - !test_and_set_bit(__QDISC_STATE_RUNNING, &q->state)) { + qdisc_run_begin(q)) { /* * This is a work-conserving queue; there are no old skbs * waiting to be sent out; and the qdisc is not running - @@ -2059,7 +2059,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, if (sch_direct_xmit(skb, q, dev, txq, root_lock)) __qdisc_run(q); else - clear_bit(__QDISC_STATE_RUNNING, &q->state); + qdisc_run_end(q); rc = NET_XMIT_SUCCESS; } else { diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index bd1892f..37b86ea 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -205,7 +205,7 @@ void __qdisc_run(struct Qdisc *q) } } - clear_bit(__QDISC_STATE_RUNNING, &q->state); + qdisc_run_end(q); } unsigned long dev_trans_start(struct net_device *dev) @@ -797,7 +797,7 @@ static bool some_qdisc_is_busy(struct net_device *dev) spin_lock_bh(root_lock); - val = (test_bit(__QDISC_STATE_RUNNING, &q->state) || + val = (qdisc_is_running(q) || test_bit(__QDISC_STATE_SCHED, &q->state)); spin_unlock_bh(root_lock); -- cgit v1.1 From 79640a4ca6955e3ebdb7038508fa7a0cd7fa5527 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 2 Jun 2010 05:09:29 -0700 Subject: net: add additional lock to qdisc to increase throughput When many cpus compete for sending frames on a given qdisc, the qdisc spinlock suffers from very high contention. The cpu owning __QDISC_STATE_RUNNING bit has same priority to acquire the lock, and cannot dequeue packets fast enough, since it must wait for this lock for each dequeued packet. One solution to this problem is to force all cpus spinning on a second lock before trying to get the main lock, when/if they see __QDISC_STATE_RUNNING already set. The owning cpu then compete with at most one other cpu for the main lock, allowing for higher dequeueing rate. Based on a previous patch from Alexander Duyck. I added the heuristic to avoid the atomic in fast path, and put the new lock far away from the cache line used by the dequeue worker. Also try to release the busylock lock as late as possible. Tests with following script gave a boost from ~50.000 pps to ~600.000 pps on a dual quad core machine (E5450 @3.00GHz), tg3 driver. (A single netperf flow can reach ~800.000 pps on this platform) for j in `seq 0 3`; do for i in `seq 0 7`; do netperf -H 192.168.0.1 -t UDP_STREAM -l 60 -N -T $i -- -m 6 & done done Signed-off-by: Eric Dumazet Acked-by: Alexander Duyck Signed-off-by: David S. Miller --- net/core/dev.c | 29 +++++++++++++++++++++++++---- net/sched/sch_generic.c | 1 + 2 files changed, 26 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/core/dev.c b/net/core/dev.c index 2733226..ffca5c10 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2040,8 +2040,18 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, struct netdev_queue *txq) { spinlock_t *root_lock = qdisc_lock(q); + bool contended = qdisc_is_running(q); int rc; + /* + * Heuristic to force contended enqueues to serialize on a + * separate lock before trying to get qdisc main lock. + * This permits __QDISC_STATE_RUNNING owner to get the lock more often + * and dequeue packets faster. + */ + if (unlikely(contended)) + spin_lock(&q->busylock); + spin_lock(root_lock); if (unlikely(test_bit(__QDISC_STATE_DEACTIVATED, &q->state))) { kfree_skb(skb); @@ -2056,19 +2066,30 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, if (!(dev->priv_flags & IFF_XMIT_DST_RELEASE)) skb_dst_force(skb); __qdisc_update_bstats(q, skb->len); - if (sch_direct_xmit(skb, q, dev, txq, root_lock)) + if (sch_direct_xmit(skb, q, dev, txq, root_lock)) { + if (unlikely(contended)) { + spin_unlock(&q->busylock); + contended = false; + } __qdisc_run(q); - else + } else qdisc_run_end(q); rc = NET_XMIT_SUCCESS; } else { skb_dst_force(skb); rc = qdisc_enqueue_root(skb, q); - qdisc_run(q); + if (qdisc_run_begin(q)) { + if (unlikely(contended)) { + spin_unlock(&q->busylock); + contended = false; + } + __qdisc_run(q); + } } spin_unlock(root_lock); - + if (unlikely(contended)) + spin_unlock(&q->busylock); return rc; } diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 37b86ea..d20fcd2 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -561,6 +561,7 @@ struct Qdisc *qdisc_alloc(struct netdev_queue *dev_queue, INIT_LIST_HEAD(&sch->list); skb_queue_head_init(&sch->q); + spin_lock_init(&sch->busylock); sch->ops = ops; sch->enqueue = ops->enqueue; sch->dequeue = ops->dequeue; -- cgit v1.1 From c2d9ba9bce8d7323ca96f239e1f505c14d6244fb Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 1 Jun 2010 06:51:19 +0000 Subject: net: CONFIG_NET_NS reduction Use read_pnet() and write_pnet() to reduce number of ifdef CONFIG_NET_NS Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv6/addrlabel.c | 6 +----- net/netfilter/nf_conntrack_core.c | 8 ++------ 2 files changed, 3 insertions(+), 11 deletions(-) (limited to 'net') diff --git a/net/ipv6/addrlabel.c b/net/ipv6/addrlabel.c index 8c4348c..f0e774c 100644 --- a/net/ipv6/addrlabel.c +++ b/net/ipv6/addrlabel.c @@ -53,11 +53,7 @@ static struct ip6addrlbl_table static inline struct net *ip6addrlbl_net(const struct ip6addrlbl_entry *lbl) { -#ifdef CONFIG_NET_NS - return lbl->lbl_net; -#else - return &init_net; -#endif + return read_pnet(&lbl->lbl_net); } /* diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c index eeeb8bc..7728898 100644 --- a/net/netfilter/nf_conntrack_core.c +++ b/net/netfilter/nf_conntrack_core.c @@ -619,9 +619,7 @@ struct nf_conn *nf_conntrack_alloc(struct net *net, u16 zone, ct->tuplehash[IP_CT_DIR_REPLY].hnnode.pprev = NULL; /* Don't set timer yet: wait for confirmation */ setup_timer(&ct->timeout, death_by_timeout, (unsigned long)ct); -#ifdef CONFIG_NET_NS - ct->ct_net = net; -#endif + write_pnet(&ct->ct_net, net); #ifdef CONFIG_NF_CONNTRACK_ZONES if (zone) { struct nf_conntrack_zone *nf_ct_zone; @@ -1363,9 +1361,7 @@ static int nf_conntrack_init_init_net(void) goto err_extend; #endif /* Set up fake conntrack: to never be deleted, not in any hashes */ -#ifdef CONFIG_NET_NS - nf_conntrack_untracked.ct_net = &init_net; -#endif + write_pnet(&nf_conntrack_untracked.ct_net, &init_net); atomic_set(&nf_conntrack_untracked.ct_general.use, 1); /* - and look it like as a confirmed connection */ set_bit(IPS_CONFIRMED_BIT, &nf_conntrack_untracked.status); -- cgit v1.1 From 614f60fa9d73a9e8fdff3df83381907fea7c5649 Mon Sep 17 00:00:00 2001 From: Scott McMillan Date: Wed, 2 Jun 2010 05:53:56 -0700 Subject: packet_mmap: expose hw packet timestamps to network packet capture utilities This patch adds a setting, PACKET_TIMESTAMP, to specify the packet timestamp source that is exported to capture utilities like tcpdump by packet_mmap. PACKET_TIMESTAMP accepts the same integer bit field as SO_TIMESTAMPING. However, only the SOF_TIMESTAMPING_SYS_HARDWARE and SOF_TIMESTAMPING_RAW_HARDWARE values are currently recognized by PACKET_TIMESTAMP. SOF_TIMESTAMPING_SYS_HARDWARE takes precedence over SOF_TIMESTAMPING_RAW_HARDWARE if both bits are set. If PACKET_TIMESTAMP is not set, a software timestamp generated inside the networking stack is used (the behavior before this setting was added). Signed-off-by: Scott McMillan Signed-off-by: David S. Miller --- net/packet/af_packet.c | 37 +++++++++++++++++++++++++++++++++++-- 1 file changed, 35 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 2078a27..9a17f28 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -83,6 +83,7 @@ #include #include #include +#include #ifdef CONFIG_INET #include @@ -202,6 +203,7 @@ struct packet_sock { unsigned int tp_hdrlen; unsigned int tp_reserve; unsigned int tp_loss:1; + unsigned int tp_tstamp; struct packet_type prot_hook ____cacheline_aligned_in_smp; }; @@ -656,6 +658,7 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, struct sk_buff *copy_skb = NULL; struct timeval tv; struct timespec ts; + struct skb_shared_hwtstamps *shhwtstamps = skb_hwtstamps(skb); if (skb->pkt_type == PACKET_LOOPBACK) goto drop; @@ -737,7 +740,13 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, h.h1->tp_snaplen = snaplen; h.h1->tp_mac = macoff; h.h1->tp_net = netoff; - if (skb->tstamp.tv64) + if ((po->tp_tstamp & SOF_TIMESTAMPING_SYS_HARDWARE) + && shhwtstamps->syststamp.tv64) + tv = ktime_to_timeval(shhwtstamps->syststamp); + else if ((po->tp_tstamp & SOF_TIMESTAMPING_RAW_HARDWARE) + && shhwtstamps->hwtstamp.tv64) + tv = ktime_to_timeval(shhwtstamps->hwtstamp); + else if (skb->tstamp.tv64) tv = ktime_to_timeval(skb->tstamp); else do_gettimeofday(&tv); @@ -750,7 +759,13 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, h.h2->tp_snaplen = snaplen; h.h2->tp_mac = macoff; h.h2->tp_net = netoff; - if (skb->tstamp.tv64) + if ((po->tp_tstamp & SOF_TIMESTAMPING_SYS_HARDWARE) + && shhwtstamps->syststamp.tv64) + ts = ktime_to_timespec(shhwtstamps->syststamp); + else if ((po->tp_tstamp & SOF_TIMESTAMPING_RAW_HARDWARE) + && shhwtstamps->hwtstamp.tv64) + ts = ktime_to_timespec(shhwtstamps->hwtstamp); + else if (skb->tstamp.tv64) ts = ktime_to_timespec(skb->tstamp); else getnstimeofday(&ts); @@ -2027,6 +2042,18 @@ packet_setsockopt(struct socket *sock, int level, int optname, char __user *optv po->has_vnet_hdr = !!val; return 0; } + case PACKET_TIMESTAMP: + { + int val; + + if (optlen != sizeof(val)) + return -EINVAL; + if (copy_from_user(&val, optval, sizeof(val))) + return -EFAULT; + + po->tp_tstamp = val; + return 0; + } default: return -ENOPROTOOPT; } @@ -2119,6 +2146,12 @@ static int packet_getsockopt(struct socket *sock, int level, int optname, val = po->tp_loss; data = &val; break; + case PACKET_TIMESTAMP: + if (len > sizeof(int)) + len = sizeof(int); + val = po->tp_tstamp; + data = &val; + break; default: return -ENOPROTOOPT; } -- cgit v1.1 From 130c0f47fdf9c533deb1136cad5136f6cb7020f0 Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Sun, 30 May 2010 17:19:53 +0000 Subject: ipconfig: send host-name in DHCP requests Normally dhclient can be configured to send the "host-name" option in DHCP requests to update the client's DNS record. However for an NFSROOT system, dhclient shall never be called (which may change the IP addr and therefore lose your root NFS mount connection). So enable updating the DNS record with kernel parameter ip=::::$HOST_NAME::dhcp Signed-off-by: Wu Fengguang Signed-off-by: David S. Miller --- net/ipv4/ipconfig.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'net') diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c index b9d84e8..3a6e1ec 100644 --- a/net/ipv4/ipconfig.c +++ b/net/ipv4/ipconfig.c @@ -665,6 +665,13 @@ ic_dhcp_init_options(u8 *options) memcpy(e, ic_req_params, sizeof(ic_req_params)); e += sizeof(ic_req_params); + if (ic_host_name_set) { + *e++ = 12; /* host-name */ + len = strlen(utsname()->nodename); + *e++ = len; + memcpy(e, utsname()->nodename, len); + e += len; + } if (*vendor_class_identifier) { printk(KERN_INFO "DHCP: sending class identifier \"%s\"\n", vendor_class_identifier); -- cgit v1.1 From 20c59de2e6b6bc74bbf714dcd4e720afe8d516cf Mon Sep 17 00:00:00 2001 From: Arnaud Ebalard Date: Tue, 1 Jun 2010 21:35:01 +0000 Subject: ipv6: Refactor update of IPv6 flowi destination address for srcrt (RH) option There are more than a dozen occurrences of following code in the IPv6 stack: if (opt && opt->srcrt) { struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt; ipv6_addr_copy(&final, &fl.fl6_dst); ipv6_addr_copy(&fl.fl6_dst, rt0->addr); final_p = &final; } Replace those with a helper. Note that the helper overrides final_p in all cases. This is ok as final_p was previously initialized to NULL when declared. Signed-off-by: Arnaud Ebalard Signed-off-by: David S. Miller --- net/dccp/ipv6.c | 30 ++++++------------------------ net/ipv6/af_inet6.c | 9 ++------- net/ipv6/datagram.c | 18 ++++-------------- net/ipv6/exthdrs.c | 24 ++++++++++++++++++++++++ net/ipv6/inet6_connection_sock.c | 9 ++------- net/ipv6/raw.c | 10 ++-------- net/ipv6/syncookies.c | 9 ++------- net/ipv6/tcp_ipv6.c | 27 ++++++--------------------- net/ipv6/udp.c | 11 +++-------- 9 files changed, 51 insertions(+), 96 deletions(-) (limited to 'net') diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c index 0916988..6e3f325 100644 --- a/net/dccp/ipv6.c +++ b/net/dccp/ipv6.c @@ -248,7 +248,7 @@ static int dccp_v6_send_response(struct sock *sk, struct request_sock *req, struct ipv6_pinfo *np = inet6_sk(sk); struct sk_buff *skb; struct ipv6_txoptions *opt = NULL; - struct in6_addr *final_p = NULL, final; + struct in6_addr *final_p, final; struct flowi fl; int err = -1; struct dst_entry *dst; @@ -265,13 +265,7 @@ static int dccp_v6_send_response(struct sock *sk, struct request_sock *req, opt = np->opt; - if (opt != NULL && opt->srcrt != NULL) { - const struct rt0_hdr *rt0 = (struct rt0_hdr *)opt->srcrt; - - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, opt, &final); err = ip6_dst_lookup(sk, &dst, &fl); if (err) @@ -545,19 +539,13 @@ static struct sock *dccp_v6_request_recv_sock(struct sock *sk, goto out_overflow; if (dst == NULL) { - struct in6_addr *final_p = NULL, final; + struct in6_addr *final_p, final; struct flowi fl; memset(&fl, 0, sizeof(fl)); fl.proto = IPPROTO_DCCP; ipv6_addr_copy(&fl.fl6_dst, &ireq6->rmt_addr); - if (opt != NULL && opt->srcrt != NULL) { - const struct rt0_hdr *rt0 = (struct rt0_hdr *)opt->srcrt; - - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, opt, &final); ipv6_addr_copy(&fl.fl6_src, &ireq6->loc_addr); fl.oif = sk->sk_bound_dev_if; fl.fl_ip_dport = inet_rsk(req)->rmt_port; @@ -885,7 +873,7 @@ static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr, struct inet_sock *inet = inet_sk(sk); struct ipv6_pinfo *np = inet6_sk(sk); struct dccp_sock *dp = dccp_sk(sk); - struct in6_addr *saddr = NULL, *final_p = NULL, final; + struct in6_addr *saddr = NULL, *final_p, final; struct flowi fl; struct dst_entry *dst; int addr_type; @@ -988,13 +976,7 @@ static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr, fl.fl_ip_sport = inet->inet_sport; security_sk_classify_flow(sk, &fl); - if (np->opt != NULL && np->opt->srcrt != NULL) { - const struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt; - - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, np->opt, &final); err = ip6_dst_lookup(sk, &dst, &fl); if (err) diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index e733942d..94b1b9c 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -651,7 +651,7 @@ int inet6_sk_rebuild_header(struct sock *sk) if (dst == NULL) { struct inet_sock *inet = inet_sk(sk); - struct in6_addr *final_p = NULL, final; + struct in6_addr *final_p, final; struct flowi fl; memset(&fl, 0, sizeof(fl)); @@ -665,12 +665,7 @@ int inet6_sk_rebuild_header(struct sock *sk) fl.fl_ip_sport = inet->inet_sport; security_sk_classify_flow(sk, &fl); - if (np->opt && np->opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *) np->opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, np->opt, &final); err = ip6_dst_lookup(sk, &dst, &fl); if (err) { diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c index 7126846..7d929a2 100644 --- a/net/ipv6/datagram.c +++ b/net/ipv6/datagram.c @@ -38,10 +38,11 @@ int ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) struct sockaddr_in6 *usin = (struct sockaddr_in6 *) uaddr; struct inet_sock *inet = inet_sk(sk); struct ipv6_pinfo *np = inet6_sk(sk); - struct in6_addr *daddr, *final_p = NULL, final; + struct in6_addr *daddr, *final_p, final; struct dst_entry *dst; struct flowi fl; struct ip6_flowlabel *flowlabel = NULL; + struct ipv6_txoptions *opt; int addr_type; int err; @@ -155,19 +156,8 @@ ipv4_connected: security_sk_classify_flow(sk, &fl); - if (flowlabel) { - if (flowlabel->opt && flowlabel->opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *) flowlabel->opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } - } else if (np->opt && np->opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + opt = flowlabel ? flowlabel->opt : np->opt; + final_p = fl6_update_dst(&fl, opt, &final); err = ip6_dst_lookup(sk, &dst, &fl); if (err) diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c index 8a659f9..853a633 100644 --- a/net/ipv6/exthdrs.c +++ b/net/ipv6/exthdrs.c @@ -874,3 +874,27 @@ struct ipv6_txoptions *ipv6_fixup_options(struct ipv6_txoptions *opt_space, return opt; } +/** + * fl6_update_dst - update flowi destination address with info given + * by srcrt option, if any. + * + * @fl: flowi for which fl6_dst is to be updated + * @opt: struct ipv6_txoptions in which to look for srcrt opt + * @orig: copy of original fl6_dst address if modified + * + * Returns NULL if no txoptions or no srcrt, otherwise returns orig + * and initial value of fl->fl6_dst set in orig + */ +struct in6_addr *fl6_update_dst(struct flowi *fl, + const struct ipv6_txoptions *opt, + struct in6_addr *orig) +{ + if (!opt || !opt->srcrt) + return NULL; + + ipv6_addr_copy(orig, &fl->fl6_dst); + ipv6_addr_copy(&fl->fl6_dst, ((struct rt0_hdr *)opt->srcrt)->addr); + return orig; +} + +EXPORT_SYMBOL_GPL(fl6_update_dst); diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c index 0c5e3c3..8a16280 100644 --- a/net/ipv6/inet6_connection_sock.c +++ b/net/ipv6/inet6_connection_sock.c @@ -185,7 +185,7 @@ int inet6_csk_xmit(struct sk_buff *skb) struct ipv6_pinfo *np = inet6_sk(sk); struct flowi fl; struct dst_entry *dst; - struct in6_addr *final_p = NULL, final; + struct in6_addr *final_p, final; memset(&fl, 0, sizeof(fl)); fl.proto = sk->sk_protocol; @@ -199,12 +199,7 @@ int inet6_csk_xmit(struct sk_buff *skb) fl.fl_ip_dport = inet->inet_dport; security_sk_classify_flow(sk, &fl); - if (np->opt && np->opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, np->opt, &final); dst = __inet6_csk_dst_check(sk, np->dst_cookie); diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 4a4dcbe..864eb8e 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -725,7 +725,7 @@ static int rawv6_sendmsg(struct kiocb *iocb, struct sock *sk, { struct ipv6_txoptions opt_space; struct sockaddr_in6 * sin6 = (struct sockaddr_in6 *) msg->msg_name; - struct in6_addr *daddr, *final_p = NULL, final; + struct in6_addr *daddr, *final_p, final; struct inet_sock *inet = inet_sk(sk); struct ipv6_pinfo *np = inet6_sk(sk); struct raw6_sock *rp = raw6_sk(sk); @@ -847,13 +847,7 @@ static int rawv6_sendmsg(struct kiocb *iocb, struct sock *sk, if (ipv6_addr_any(&fl.fl6_src) && !ipv6_addr_any(&np->saddr)) ipv6_addr_copy(&fl.fl6_src, &np->saddr); - /* merge ip6_build_xmit from ip6_output */ - if (opt && opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, opt, &final); if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst)) fl.oif = np->mcast_oif; diff --git a/net/ipv6/syncookies.c b/net/ipv6/syncookies.c index 34d1f06..1238370 100644 --- a/net/ipv6/syncookies.c +++ b/net/ipv6/syncookies.c @@ -240,17 +240,12 @@ struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb) * me if there is a preferred way. */ { - struct in6_addr *final_p = NULL, final; + struct in6_addr *final_p, final; struct flowi fl; memset(&fl, 0, sizeof(fl)); fl.proto = IPPROTO_TCP; ipv6_addr_copy(&fl.fl6_dst, &ireq6->rmt_addr); - if (np->opt && np->opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *) np->opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, np->opt, &final); ipv6_addr_copy(&fl.fl6_src, &ireq6->loc_addr); fl.oif = sk->sk_bound_dev_if; fl.mark = sk->sk_mark; diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 2b7c3a1..e487080 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -129,7 +129,7 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, struct inet_connection_sock *icsk = inet_csk(sk); struct ipv6_pinfo *np = inet6_sk(sk); struct tcp_sock *tp = tcp_sk(sk); - struct in6_addr *saddr = NULL, *final_p = NULL, final; + struct in6_addr *saddr = NULL, *final_p, final; struct flowi fl; struct dst_entry *dst; int addr_type; @@ -250,12 +250,7 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, fl.fl_ip_dport = usin->sin6_port; fl.fl_ip_sport = inet->inet_sport; - if (np->opt && np->opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, np->opt, &final); security_sk_classify_flow(sk, &fl); @@ -477,7 +472,7 @@ static int tcp_v6_send_synack(struct sock *sk, struct request_sock *req, struct ipv6_pinfo *np = inet6_sk(sk); struct sk_buff * skb; struct ipv6_txoptions *opt = NULL; - struct in6_addr * final_p = NULL, final; + struct in6_addr * final_p, final; struct flowi fl; struct dst_entry *dst; int err = -1; @@ -494,12 +489,7 @@ static int tcp_v6_send_synack(struct sock *sk, struct request_sock *req, security_req_classify_flow(req, &fl); opt = np->opt; - if (opt && opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, opt, &final); err = ip6_dst_lookup(sk, &dst, &fl); if (err) @@ -1392,18 +1382,13 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, goto out_overflow; if (dst == NULL) { - struct in6_addr *final_p = NULL, final; + struct in6_addr *final_p, final; struct flowi fl; memset(&fl, 0, sizeof(fl)); fl.proto = IPPROTO_TCP; ipv6_addr_copy(&fl.fl6_dst, &treq->rmt_addr); - if (opt && opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } + final_p = fl6_update_dst(&fl, opt, &final); ipv6_addr_copy(&fl.fl6_src, &treq->loc_addr); fl.oif = sk->sk_bound_dev_if; fl.mark = sk->sk_mark; diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index 3048f90..4aea57d 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -929,7 +929,7 @@ int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk, struct inet_sock *inet = inet_sk(sk); struct ipv6_pinfo *np = inet6_sk(sk); struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *) msg->msg_name; - struct in6_addr *daddr, *final_p = NULL, final; + struct in6_addr *daddr, *final_p, final; struct ipv6_txoptions *opt = NULL; struct ip6_flowlabel *flowlabel = NULL; struct flowi fl; @@ -1099,14 +1099,9 @@ do_udp_sendmsg: ipv6_addr_copy(&fl.fl6_src, &np->saddr); fl.fl_ip_sport = inet->inet_sport; - /* merge ip6_build_xmit from ip6_output */ - if (opt && opt->srcrt) { - struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt; - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; + final_p = fl6_update_dst(&fl, opt, &final); + if (final_p) connected = 0; - } if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst)) { fl.oif = np->mcast_oif; -- cgit v1.1 From ab95bfe01f9872459c8678572ccadbf646badad0 Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Tue, 1 Jun 2010 21:52:08 +0000 Subject: net: replace hooks in __netif_receive_skb V5 What this patch does is it removes two receive frame hooks (for bridge and for macvlan) from __netif_receive_skb. These are replaced them with a single hook for both. It only supports one hook per device because it makes no sense to do bridging and macvlan on the same device. Then a network driver (of virtual netdev like macvlan or bridge) can register an rx_handler for needed net device. Signed-off-by: Jiri Pirko Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/bridge/br.c | 2 - net/bridge/br_if.c | 8 ++++ net/bridge/br_input.c | 12 +++-- net/bridge/br_private.h | 3 +- net/core/dev.c | 119 ++++++++++++++++++++++-------------------------- 5 files changed, 73 insertions(+), 71 deletions(-) (limited to 'net') diff --git a/net/bridge/br.c b/net/bridge/br.c index 76357b5..c8436fa 100644 --- a/net/bridge/br.c +++ b/net/bridge/br.c @@ -63,7 +63,6 @@ static int __init br_init(void) goto err_out4; brioctl_set(br_ioctl_deviceless_stub); - br_handle_frame_hook = br_handle_frame; #if defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE) br_fdb_test_addr_hook = br_fdb_test_addr; @@ -100,7 +99,6 @@ static void __exit br_deinit(void) br_fdb_test_addr_hook = NULL; #endif - br_handle_frame_hook = NULL; br_fdb_fini(); } diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c index 18b245e..d924234 100644 --- a/net/bridge/br_if.c +++ b/net/bridge/br_if.c @@ -147,6 +147,7 @@ static void del_nbp(struct net_bridge_port *p) list_del_rcu(&p->list); + netdev_rx_handler_unregister(dev); rcu_assign_pointer(dev->br_port, NULL); br_multicast_del_port(p); @@ -429,6 +430,11 @@ int br_add_if(struct net_bridge *br, struct net_device *dev) goto err2; rcu_assign_pointer(dev->br_port, p); + + err = netdev_rx_handler_register(dev, br_handle_frame); + if (err) + goto err3; + dev_disable_lro(dev); list_add_rcu(&p->list, &br->port_list); @@ -451,6 +457,8 @@ int br_add_if(struct net_bridge *br, struct net_device *dev) br_netpoll_enable(br, dev); return 0; +err3: + rcu_assign_pointer(dev->br_port, NULL); err2: br_fdb_delete_by_port(br, p, 1); err1: diff --git a/net/bridge/br_input.c b/net/bridge/br_input.c index d36e700..99647d8 100644 --- a/net/bridge/br_input.c +++ b/net/bridge/br_input.c @@ -131,15 +131,19 @@ static inline int is_link_local(const unsigned char *dest) } /* - * Called via br_handle_frame_hook. * Return NULL if skb is handled - * note: already called with rcu_read_lock (preempt_disabled) + * note: already called with rcu_read_lock (preempt_disabled) from + * netif_receive_skb */ -struct sk_buff *br_handle_frame(struct net_bridge_port *p, struct sk_buff *skb) +struct sk_buff *br_handle_frame(struct sk_buff *skb) { + struct net_bridge_port *p; const unsigned char *dest = eth_hdr(skb)->h_dest; int (*rhook)(struct sk_buff *skb); + if (skb->pkt_type == PACKET_LOOPBACK) + return skb; + if (!is_valid_ether_addr(eth_hdr(skb)->h_source)) goto drop; @@ -147,6 +151,8 @@ struct sk_buff *br_handle_frame(struct net_bridge_port *p, struct sk_buff *skb) if (!skb) return NULL; + p = rcu_dereference(skb->dev->br_port); + if (unlikely(is_link_local(dest))) { /* Pause frames shouldn't be passed up by driver anyway */ if (skb->protocol == htons(ETH_P_PAUSE)) diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h index 0f4a74b..c83519b 100644 --- a/net/bridge/br_private.h +++ b/net/bridge/br_private.h @@ -331,8 +331,7 @@ extern void br_features_recompute(struct net_bridge *br); /* br_input.c */ extern int br_handle_frame_finish(struct sk_buff *skb); -extern struct sk_buff *br_handle_frame(struct net_bridge_port *p, - struct sk_buff *skb); +extern struct sk_buff *br_handle_frame(struct sk_buff *skb); /* br_ioctl.c */ extern int br_dev_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); diff --git a/net/core/dev.c b/net/core/dev.c index ffca5c10..ec01a59 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2604,70 +2604,14 @@ static inline int deliver_skb(struct sk_buff *skb, return pt_prev->func(skb, skb->dev, pt_prev, orig_dev); } -#if defined(CONFIG_BRIDGE) || defined (CONFIG_BRIDGE_MODULE) - -#if defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE) +#if (defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)) && \ + (defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE)) /* This hook is defined here for ATM LANE */ int (*br_fdb_test_addr_hook)(struct net_device *dev, unsigned char *addr) __read_mostly; EXPORT_SYMBOL_GPL(br_fdb_test_addr_hook); #endif -/* - * If bridge module is loaded call bridging hook. - * returns NULL if packet was consumed. - */ -struct sk_buff *(*br_handle_frame_hook)(struct net_bridge_port *p, - struct sk_buff *skb) __read_mostly; -EXPORT_SYMBOL_GPL(br_handle_frame_hook); - -static inline struct sk_buff *handle_bridge(struct sk_buff *skb, - struct packet_type **pt_prev, int *ret, - struct net_device *orig_dev) -{ - struct net_bridge_port *port; - - if (skb->pkt_type == PACKET_LOOPBACK || - (port = rcu_dereference(skb->dev->br_port)) == NULL) - return skb; - - if (*pt_prev) { - *ret = deliver_skb(skb, *pt_prev, orig_dev); - *pt_prev = NULL; - } - - return br_handle_frame_hook(port, skb); -} -#else -#define handle_bridge(skb, pt_prev, ret, orig_dev) (skb) -#endif - -#if defined(CONFIG_MACVLAN) || defined(CONFIG_MACVLAN_MODULE) -struct sk_buff *(*macvlan_handle_frame_hook)(struct macvlan_port *p, - struct sk_buff *skb) __read_mostly; -EXPORT_SYMBOL_GPL(macvlan_handle_frame_hook); - -static inline struct sk_buff *handle_macvlan(struct sk_buff *skb, - struct packet_type **pt_prev, - int *ret, - struct net_device *orig_dev) -{ - struct macvlan_port *port; - - port = rcu_dereference(skb->dev->macvlan_port); - if (!port) - return skb; - - if (*pt_prev) { - *ret = deliver_skb(skb, *pt_prev, orig_dev); - *pt_prev = NULL; - } - return macvlan_handle_frame_hook(port, skb); -} -#else -#define handle_macvlan(skb, pt_prev, ret, orig_dev) (skb) -#endif - #ifdef CONFIG_NET_CLS_ACT /* TODO: Maybe we should just force sch_ingress to be compiled in * when CONFIG_NET_CLS_ACT is? otherwise some useless instructions @@ -2763,6 +2707,47 @@ void netif_nit_deliver(struct sk_buff *skb) rcu_read_unlock(); } +/** + * netdev_rx_handler_register - register receive handler + * @dev: device to register a handler for + * @rx_handler: receive handler to register + * + * Register a receive hander for a device. This handler will then be + * called from __netif_receive_skb. A negative errno code is returned + * on a failure. + * + * The caller must hold the rtnl_mutex. + */ +int netdev_rx_handler_register(struct net_device *dev, + rx_handler_func_t *rx_handler) +{ + ASSERT_RTNL(); + + if (dev->rx_handler) + return -EBUSY; + + rcu_assign_pointer(dev->rx_handler, rx_handler); + + return 0; +} +EXPORT_SYMBOL_GPL(netdev_rx_handler_register); + +/** + * netdev_rx_handler_unregister - unregister receive handler + * @dev: device to unregister a handler from + * + * Unregister a receive hander from a device. + * + * The caller must hold the rtnl_mutex. + */ +void netdev_rx_handler_unregister(struct net_device *dev) +{ + + ASSERT_RTNL(); + rcu_assign_pointer(dev->rx_handler, NULL); +} +EXPORT_SYMBOL_GPL(netdev_rx_handler_unregister); + static inline void skb_bond_set_mac_by_master(struct sk_buff *skb, struct net_device *master) { @@ -2815,6 +2800,7 @@ EXPORT_SYMBOL(__skb_bond_should_drop); static int __netif_receive_skb(struct sk_buff *skb) { struct packet_type *ptype, *pt_prev; + rx_handler_func_t *rx_handler; struct net_device *orig_dev; struct net_device *master; struct net_device *null_or_orig; @@ -2877,12 +2863,17 @@ static int __netif_receive_skb(struct sk_buff *skb) ncls: #endif - skb = handle_bridge(skb, &pt_prev, &ret, orig_dev); - if (!skb) - goto out; - skb = handle_macvlan(skb, &pt_prev, &ret, orig_dev); - if (!skb) - goto out; + /* Handle special case of bridge or macvlan */ + rx_handler = rcu_dereference(skb->dev->rx_handler); + if (rx_handler) { + if (pt_prev) { + ret = deliver_skb(skb, pt_prev, orig_dev); + pt_prev = NULL; + } + skb = rx_handler(skb); + if (!skb) + goto out; + } /* * Make sure frames received on VLAN interfaces stacked on -- cgit v1.1 From ec8aa669b8393b6789b1954d587c63264af7ff99 Mon Sep 17 00:00:00 2001 From: Felix Fietkau Date: Thu, 13 May 2010 16:48:03 +0200 Subject: mac80211: add the minstrel_ht rate control algorithm Signed-off-by: Felix Fietkau Signed-off-by: John W. Linville --- net/mac80211/Kconfig | 7 + net/mac80211/Makefile | 4 + net/mac80211/main.c | 7 + net/mac80211/rate.h | 13 + net/mac80211/rc80211_minstrel_ht.c | 824 +++++++++++++++++++++++++++++ net/mac80211/rc80211_minstrel_ht.h | 128 +++++ net/mac80211/rc80211_minstrel_ht_debugfs.c | 120 +++++ 7 files changed, 1103 insertions(+) create mode 100644 net/mac80211/rc80211_minstrel_ht.c create mode 100644 net/mac80211/rc80211_minstrel_ht.h create mode 100644 net/mac80211/rc80211_minstrel_ht_debugfs.c (limited to 'net') diff --git a/net/mac80211/Kconfig b/net/mac80211/Kconfig index 8a91f6c..83eec7a 100644 --- a/net/mac80211/Kconfig +++ b/net/mac80211/Kconfig @@ -33,6 +33,13 @@ config MAC80211_RC_MINSTREL ---help--- This option enables the 'minstrel' TX rate control algorithm +config MAC80211_RC_MINSTREL_HT + bool "Minstrel 802.11n support" if EMBEDDED + depends on MAC80211_RC_MINSTREL + default y + ---help--- + This option enables the 'minstrel_ht' TX rate control algorithm + choice prompt "Default rate control algorithm" depends on MAC80211_HAS_RC diff --git a/net/mac80211/Makefile b/net/mac80211/Makefile index 84b48ba..fdb54e6 100644 --- a/net/mac80211/Makefile +++ b/net/mac80211/Makefile @@ -51,7 +51,11 @@ rc80211_pid-$(CONFIG_MAC80211_DEBUGFS) += rc80211_pid_debugfs.o rc80211_minstrel-y := rc80211_minstrel.o rc80211_minstrel-$(CONFIG_MAC80211_DEBUGFS) += rc80211_minstrel_debugfs.o +rc80211_minstrel_ht-y := rc80211_minstrel_ht.o +rc80211_minstrel_ht-$(CONFIG_MAC80211_DEBUGFS) += rc80211_minstrel_ht_debugfs.o + mac80211-$(CONFIG_MAC80211_RC_PID) += $(rc80211_pid-y) mac80211-$(CONFIG_MAC80211_RC_MINSTREL) += $(rc80211_minstrel-y) +mac80211-$(CONFIG_MAC80211_RC_MINSTREL_HT) += $(rc80211_minstrel_ht-y) ccflags-y += -D__CHECK_ENDIAN__ diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 22a384d..c8548e61 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -704,6 +704,10 @@ static int __init ieee80211_init(void) if (ret) return ret; + ret = rc80211_minstrel_ht_init(); + if (ret) + goto err_minstrel; + ret = rc80211_pid_init(); if (ret) goto err_pid; @@ -716,6 +720,8 @@ static int __init ieee80211_init(void) err_netdev: rc80211_pid_exit(); err_pid: + rc80211_minstrel_ht_exit(); + err_minstrel: rc80211_minstrel_exit(); return ret; @@ -724,6 +730,7 @@ static int __init ieee80211_init(void) static void __exit ieee80211_exit(void) { rc80211_pid_exit(); + rc80211_minstrel_ht_exit(); rc80211_minstrel_exit(); /* diff --git a/net/mac80211/rate.h b/net/mac80211/rate.h index 065a9619..168427b 100644 --- a/net/mac80211/rate.h +++ b/net/mac80211/rate.h @@ -147,5 +147,18 @@ static inline void rc80211_minstrel_exit(void) } #endif +#ifdef CONFIG_MAC80211_RC_MINSTREL_HT +extern int rc80211_minstrel_ht_init(void); +extern void rc80211_minstrel_ht_exit(void); +#else +static inline int rc80211_minstrel_ht_init(void) +{ + return 0; +} +static inline void rc80211_minstrel_ht_exit(void) +{ +} +#endif + #endif /* IEEE80211_RATE_H */ diff --git a/net/mac80211/rc80211_minstrel_ht.c b/net/mac80211/rc80211_minstrel_ht.c new file mode 100644 index 0000000..c23f082 --- /dev/null +++ b/net/mac80211/rc80211_minstrel_ht.c @@ -0,0 +1,824 @@ +/* + * Copyright (C) 2010 Felix Fietkau + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include +#include +#include +#include +#include +#include +#include +#include "rate.h" +#include "rc80211_minstrel.h" +#include "rc80211_minstrel_ht.h" + +#define AVG_PKT_SIZE 1200 +#define SAMPLE_COLUMNS 10 +#define EWMA_LEVEL 75 + +/* Number of bits for an average sized packet */ +#define MCS_NBITS (AVG_PKT_SIZE << 3) + +/* Number of symbols for a packet with (bps) bits per symbol */ +#define MCS_NSYMS(bps) ((MCS_NBITS + (bps) - 1) / (bps)) + +/* Transmission time for a packet containing (syms) symbols */ +#define MCS_SYMBOL_TIME(sgi, syms) \ + (sgi ? \ + ((syms) * 18 + 4) / 5 : /* syms * 3.6 us */ \ + (syms) << 2 /* syms * 4 us */ \ + ) + +/* Transmit duration for the raw data part of an average sized packet */ +#define MCS_DURATION(streams, sgi, bps) MCS_SYMBOL_TIME(sgi, MCS_NSYMS((streams) * (bps))) + +/* MCS rate information for an MCS group */ +#define MCS_GROUP(_streams, _sgi, _ht40) { \ + .streams = _streams, \ + .flags = \ + (_sgi ? IEEE80211_TX_RC_SHORT_GI : 0) | \ + (_ht40 ? IEEE80211_TX_RC_40_MHZ_WIDTH : 0), \ + .duration = { \ + MCS_DURATION(_streams, _sgi, _ht40 ? 54 : 26), \ + MCS_DURATION(_streams, _sgi, _ht40 ? 108 : 52), \ + MCS_DURATION(_streams, _sgi, _ht40 ? 162 : 78), \ + MCS_DURATION(_streams, _sgi, _ht40 ? 216 : 104), \ + MCS_DURATION(_streams, _sgi, _ht40 ? 324 : 156), \ + MCS_DURATION(_streams, _sgi, _ht40 ? 432 : 208), \ + MCS_DURATION(_streams, _sgi, _ht40 ? 486 : 234), \ + MCS_DURATION(_streams, _sgi, _ht40 ? 540 : 260) \ + } \ +} + +/* + * To enable sufficiently targeted rate sampling, MCS rates are divided into + * groups, based on the number of streams and flags (HT40, SGI) that they + * use. + */ +const struct mcs_group minstrel_mcs_groups[] = { + MCS_GROUP(1, 0, 0), + MCS_GROUP(2, 0, 0), +#if MINSTREL_MAX_STREAMS >= 3 + MCS_GROUP(3, 0, 0), +#endif + + MCS_GROUP(1, 1, 0), + MCS_GROUP(2, 1, 0), +#if MINSTREL_MAX_STREAMS >= 3 + MCS_GROUP(3, 1, 0), +#endif + + MCS_GROUP(1, 0, 1), + MCS_GROUP(2, 0, 1), +#if MINSTREL_MAX_STREAMS >= 3 + MCS_GROUP(3, 0, 1), +#endif + + MCS_GROUP(1, 1, 1), + MCS_GROUP(2, 1, 1), +#if MINSTREL_MAX_STREAMS >= 3 + MCS_GROUP(3, 1, 1), +#endif +}; + +static u8 sample_table[SAMPLE_COLUMNS][MCS_GROUP_RATES]; + +/* + * Perform EWMA (Exponentially Weighted Moving Average) calculation + */ +static int +minstrel_ewma(int old, int new, int weight) +{ + return (new * (100 - weight) + old * weight) / 100; +} + +/* + * Look up an MCS group index based on mac80211 rate information + */ +static int +minstrel_ht_get_group_idx(struct ieee80211_tx_rate *rate) +{ + int streams = (rate->idx / MCS_GROUP_RATES) + 1; + u32 flags = IEEE80211_TX_RC_SHORT_GI | IEEE80211_TX_RC_40_MHZ_WIDTH; + int i; + + for (i = 0; i < ARRAY_SIZE(minstrel_mcs_groups); i++) { + if (minstrel_mcs_groups[i].streams != streams) + continue; + if (minstrel_mcs_groups[i].flags != (rate->flags & flags)) + continue; + + return i; + } + + WARN_ON(1); + return 0; +} + +static inline struct minstrel_rate_stats * +minstrel_get_ratestats(struct minstrel_ht_sta *mi, int index) +{ + return &mi->groups[index / MCS_GROUP_RATES].rates[index % MCS_GROUP_RATES]; +} + + +/* + * Recalculate success probabilities and counters for a rate using EWMA + */ +static void +minstrel_calc_rate_ewma(struct minstrel_priv *mp, struct minstrel_rate_stats *mr) +{ + if (unlikely(mr->attempts > 0)) { + mr->sample_skipped = 0; + mr->cur_prob = MINSTREL_FRAC(mr->success, mr->attempts); + if (!mr->att_hist) + mr->probability = mr->cur_prob; + else + mr->probability = minstrel_ewma(mr->probability, + mr->cur_prob, EWMA_LEVEL); + mr->att_hist += mr->attempts; + mr->succ_hist += mr->success; + } else { + mr->sample_skipped++; + } + mr->last_success = mr->success; + mr->last_attempts = mr->attempts; + mr->success = 0; + mr->attempts = 0; +} + +/* + * Calculate throughput based on the average A-MPDU length, taking into account + * the expected number of retransmissions and their expected length + */ +static void +minstrel_ht_calc_tp(struct minstrel_priv *mp, struct minstrel_ht_sta *mi, + int group, int rate) +{ + struct minstrel_rate_stats *mr; + unsigned int usecs; + + mr = &mi->groups[group].rates[rate]; + + if (mr->probability < MINSTREL_FRAC(1, 10)) { + mr->cur_tp = 0; + return; + } + + usecs = mi->overhead / MINSTREL_TRUNC(mi->avg_ampdu_len); + usecs += minstrel_mcs_groups[group].duration[rate]; + mr->cur_tp = MINSTREL_TRUNC((1000000 / usecs) * mr->probability); +} + +/* + * Update rate statistics and select new primary rates + * + * Rules for rate selection: + * - max_prob_rate must use only one stream, as a tradeoff between delivery + * probability and throughput during strong fluctuations + * - as long as the max prob rate has a probability of more than 3/4, pick + * higher throughput rates, even if the probablity is a bit lower + */ +static void +minstrel_ht_update_stats(struct minstrel_priv *mp, struct minstrel_ht_sta *mi) +{ + struct minstrel_mcs_group_data *mg; + struct minstrel_rate_stats *mr; + int cur_prob, cur_prob_tp, cur_tp, cur_tp2; + int group, i, index; + + if (mi->ampdu_packets > 0) { + mi->avg_ampdu_len = minstrel_ewma(mi->avg_ampdu_len, + MINSTREL_FRAC(mi->ampdu_len, mi->ampdu_packets), EWMA_LEVEL); + mi->ampdu_len = 0; + mi->ampdu_packets = 0; + } + + mi->sample_slow = 0; + mi->sample_count = 0; + mi->max_tp_rate = 0; + mi->max_tp_rate2 = 0; + mi->max_prob_rate = 0; + + for (group = 0; group < ARRAY_SIZE(minstrel_mcs_groups); group++) { + cur_prob = 0; + cur_prob_tp = 0; + cur_tp = 0; + cur_tp2 = 0; + + mg = &mi->groups[group]; + if (!mg->supported) + continue; + + mg->max_tp_rate = 0; + mg->max_tp_rate2 = 0; + mg->max_prob_rate = 0; + mi->sample_count++; + + for (i = 0; i < MCS_GROUP_RATES; i++) { + if (!(mg->supported & BIT(i))) + continue; + + mr = &mg->rates[i]; + mr->retry_updated = false; + index = MCS_GROUP_RATES * group + i; + minstrel_calc_rate_ewma(mp, mr); + minstrel_ht_calc_tp(mp, mi, group, i); + + if (!mr->cur_tp) + continue; + + /* ignore the lowest rate of each single-stream group */ + if (!i && minstrel_mcs_groups[group].streams == 1) + continue; + + if ((mr->cur_tp > cur_prob_tp && mr->probability > + MINSTREL_FRAC(3, 4)) || mr->probability > cur_prob) { + mg->max_prob_rate = index; + cur_prob = mr->probability; + } + + if (mr->cur_tp > cur_tp) { + swap(index, mg->max_tp_rate); + cur_tp = mr->cur_tp; + mr = minstrel_get_ratestats(mi, index); + } + + if (index >= mg->max_tp_rate) + continue; + + if (mr->cur_tp > cur_tp2) { + mg->max_tp_rate2 = index; + cur_tp2 = mr->cur_tp; + } + } + } + + /* try to sample up to half of the availble rates during each interval */ + mi->sample_count *= 4; + + cur_prob = 0; + cur_prob_tp = 0; + cur_tp = 0; + cur_tp2 = 0; + for (group = 0; group < ARRAY_SIZE(minstrel_mcs_groups); group++) { + mg = &mi->groups[group]; + if (!mg->supported) + continue; + + mr = minstrel_get_ratestats(mi, mg->max_prob_rate); + if (cur_prob_tp < mr->cur_tp && + minstrel_mcs_groups[group].streams == 1) { + mi->max_prob_rate = mg->max_prob_rate; + cur_prob = mr->cur_prob; + } + + mr = minstrel_get_ratestats(mi, mg->max_tp_rate); + if (cur_tp < mr->cur_tp) { + mi->max_tp_rate = mg->max_tp_rate; + cur_tp = mr->cur_tp; + } + + mr = minstrel_get_ratestats(mi, mg->max_tp_rate2); + if (cur_tp2 < mr->cur_tp) { + mi->max_tp_rate2 = mg->max_tp_rate2; + cur_tp2 = mr->cur_tp; + } + } + + mi->stats_update = jiffies; +} + +static bool +minstrel_ht_txstat_valid(struct ieee80211_tx_rate *rate) +{ + if (!rate->count) + return false; + + if (rate->idx < 0) + return false; + + return !!(rate->flags & IEEE80211_TX_RC_MCS); +} + +static void +minstrel_next_sample_idx(struct minstrel_ht_sta *mi) +{ + struct minstrel_mcs_group_data *mg; + + for (;;) { + mi->sample_group++; + mi->sample_group %= ARRAY_SIZE(minstrel_mcs_groups); + mg = &mi->groups[mi->sample_group]; + + if (!mg->supported) + continue; + + if (++mg->index >= MCS_GROUP_RATES) { + mg->index = 0; + if (++mg->column >= ARRAY_SIZE(sample_table)) + mg->column = 0; + } + break; + } +} + +static void +minstrel_downgrade_rate(struct minstrel_ht_sta *mi, int *idx, bool primary) +{ + int group, orig_group; + + orig_group = group = *idx / MCS_GROUP_RATES; + while (group > 0) { + group--; + + if (!mi->groups[group].supported) + continue; + + if (minstrel_mcs_groups[group].streams > + minstrel_mcs_groups[orig_group].streams) + continue; + + if (primary) + *idx = mi->groups[group].max_tp_rate; + else + *idx = mi->groups[group].max_tp_rate2; + break; + } +} + +static void +minstrel_aggr_check(struct minstrel_priv *mp, struct ieee80211_sta *pubsta, struct sk_buff *skb) +{ + struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data; + struct sta_info *sta = container_of(pubsta, struct sta_info, sta); + u16 tid; + + if (unlikely(!ieee80211_is_data_qos(hdr->frame_control))) + return; + + if (unlikely(skb->protocol == cpu_to_be16(ETH_P_PAE))) + return; + + tid = *ieee80211_get_qos_ctl(hdr) & IEEE80211_QOS_CTL_TID_MASK; + if (likely(sta->ampdu_mlme.tid_state_tx[tid] != HT_AGG_STATE_IDLE)) + return; + + ieee80211_start_tx_ba_session(pubsta, tid); +} + +static void +minstrel_ht_tx_status(void *priv, struct ieee80211_supported_band *sband, + struct ieee80211_sta *sta, void *priv_sta, + struct sk_buff *skb) +{ + struct minstrel_ht_sta_priv *msp = priv_sta; + struct minstrel_ht_sta *mi = &msp->ht; + struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); + struct ieee80211_tx_rate *ar = info->status.rates; + struct minstrel_rate_stats *rate, *rate2; + struct minstrel_priv *mp = priv; + bool last = false; + int group; + int i = 0; + + if (!msp->is_ht) + return mac80211_minstrel.tx_status(priv, sband, sta, &msp->legacy, skb); + + /* This packet was aggregated but doesn't carry status info */ + if ((info->flags & IEEE80211_TX_CTL_AMPDU) && + !(info->flags & IEEE80211_TX_STAT_AMPDU)) + return; + + if (!info->status.ampdu_len) { + info->status.ampdu_ack_len = 1; + info->status.ampdu_len = 1; + } + + mi->ampdu_packets++; + mi->ampdu_len += info->status.ampdu_len; + + if (!mi->sample_wait && !mi->sample_tries && mi->sample_count > 0) { + mi->sample_wait = 4 + 2 * MINSTREL_TRUNC(mi->avg_ampdu_len); + mi->sample_tries = 3; + mi->sample_count--; + } + + if (info->flags & IEEE80211_TX_CTL_RATE_CTRL_PROBE) { + mi->sample_packets += info->status.ampdu_len; + minstrel_next_sample_idx(mi); + } + + for (i = 0; !last; i++) { + last = (i == IEEE80211_TX_MAX_RATES - 1) || + !minstrel_ht_txstat_valid(&ar[i + 1]); + + if (!minstrel_ht_txstat_valid(&ar[i])) + break; + + group = minstrel_ht_get_group_idx(&ar[i]); + rate = &mi->groups[group].rates[ar[i].idx % 8]; + + if (last && (info->flags & IEEE80211_TX_STAT_ACK)) + rate->success += info->status.ampdu_ack_len; + + rate->attempts += ar[i].count * info->status.ampdu_len; + } + + /* + * check for sudden death of spatial multiplexing, + * downgrade to a lower number of streams if necessary. + */ + rate = minstrel_get_ratestats(mi, mi->max_tp_rate); + if (rate->attempts > 30 && + MINSTREL_FRAC(rate->success, rate->attempts) < + MINSTREL_FRAC(20, 100)) + minstrel_downgrade_rate(mi, &mi->max_tp_rate, true); + + rate2 = minstrel_get_ratestats(mi, mi->max_tp_rate2); + if (rate->attempts > 30 && + MINSTREL_FRAC(rate->success, rate->attempts) < + MINSTREL_FRAC(20, 100)) + minstrel_downgrade_rate(mi, &mi->max_tp_rate2, false); + + if (time_after(jiffies, mi->stats_update + (mp->update_interval / 2 * HZ) / 1000)) { + minstrel_ht_update_stats(mp, mi); + minstrel_aggr_check(mp, sta, skb); + } +} + +static void +minstrel_calc_retransmit(struct minstrel_priv *mp, struct minstrel_ht_sta *mi, + int index) +{ + struct minstrel_rate_stats *mr; + const struct mcs_group *group; + unsigned int tx_time, tx_time_rtscts, tx_time_data; + unsigned int cw = mp->cw_min; + unsigned int t_slot = 9; /* FIXME */ + unsigned int ampdu_len = MINSTREL_TRUNC(mi->avg_ampdu_len); + + mr = minstrel_get_ratestats(mi, index); + if (mr->probability < MINSTREL_FRAC(1, 10)) { + mr->retry_count = 1; + mr->retry_count_rtscts = 1; + return; + } + + mr->retry_count = 2; + mr->retry_count_rtscts = 2; + mr->retry_updated = true; + + group = &minstrel_mcs_groups[index / MCS_GROUP_RATES]; + tx_time_data = group->duration[index % MCS_GROUP_RATES] * ampdu_len; + tx_time = 2 * (t_slot + mi->overhead + tx_time_data); + tx_time_rtscts = 2 * (t_slot + mi->overhead_rtscts + tx_time_data); + do { + cw = (cw << 1) | 1; + cw = min(cw, mp->cw_max); + tx_time += cw + t_slot + mi->overhead; + tx_time_rtscts += cw + t_slot + mi->overhead_rtscts; + if (tx_time_rtscts < mp->segment_size) + mr->retry_count_rtscts++; + } while ((tx_time < mp->segment_size) && + (++mr->retry_count < mp->max_retry)); +} + + +static void +minstrel_ht_set_rate(struct minstrel_priv *mp, struct minstrel_ht_sta *mi, + struct ieee80211_tx_rate *rate, int index, + struct ieee80211_tx_rate_control *txrc, + bool sample, bool rtscts) +{ + const struct mcs_group *group = &minstrel_mcs_groups[index / MCS_GROUP_RATES]; + struct minstrel_rate_stats *mr; + + mr = minstrel_get_ratestats(mi, index); + if (!mr->retry_updated) + minstrel_calc_retransmit(mp, mi, index); + + if (mr->probability < MINSTREL_FRAC(20, 100)) + rate->count = 2; + else if (rtscts) + rate->count = mr->retry_count_rtscts; + else + rate->count = mr->retry_count; + + rate->flags = IEEE80211_TX_RC_MCS | group->flags; + if (txrc->short_preamble) + rate->flags |= IEEE80211_TX_RC_USE_SHORT_PREAMBLE; + if (txrc->rts || rtscts) + rate->flags |= IEEE80211_TX_RC_USE_RTS_CTS; + rate->idx = index % MCS_GROUP_RATES + (group->streams - 1) * MCS_GROUP_RATES; +} + +static inline int +minstrel_get_duration(int index) +{ + const struct mcs_group *group = &minstrel_mcs_groups[index / MCS_GROUP_RATES]; + return group->duration[index % MCS_GROUP_RATES]; +} + +static int +minstrel_get_sample_rate(struct minstrel_priv *mp, struct minstrel_ht_sta *mi) +{ + struct minstrel_rate_stats *mr; + struct minstrel_mcs_group_data *mg; + int sample_idx = 0; + + if (mi->sample_wait > 0) { + mi->sample_wait--; + return -1; + } + + if (!mi->sample_tries) + return -1; + + mi->sample_tries--; + mg = &mi->groups[mi->sample_group]; + sample_idx = sample_table[mg->column][mg->index]; + mr = &mg->rates[sample_idx]; + sample_idx += mi->sample_group * MCS_GROUP_RATES; + + /* + * When not using MRR, do not sample if the probability is already + * higher than 95% to avoid wasting airtime + */ + if (!mp->has_mrr && (mr->probability > MINSTREL_FRAC(95, 100))) + goto next; + + /* + * Make sure that lower rates get sampled only occasionally, + * if the link is working perfectly. + */ + if (minstrel_get_duration(sample_idx) > + minstrel_get_duration(mi->max_tp_rate)) { + if (mr->sample_skipped < 10) + goto next; + + if (mi->sample_slow++ > 2) + goto next; + } + + return sample_idx; + +next: + minstrel_next_sample_idx(mi); + return -1; +} + +static void +minstrel_ht_get_rate(void *priv, struct ieee80211_sta *sta, void *priv_sta, + struct ieee80211_tx_rate_control *txrc) +{ + struct ieee80211_tx_info *info = IEEE80211_SKB_CB(txrc->skb); + struct ieee80211_tx_rate *ar = info->status.rates; + struct minstrel_ht_sta_priv *msp = priv_sta; + struct minstrel_ht_sta *mi = &msp->ht; + struct minstrel_priv *mp = priv; + int sample_idx; + + if (rate_control_send_low(sta, priv_sta, txrc)) + return; + + if (!msp->is_ht) + return mac80211_minstrel.get_rate(priv, sta, &msp->legacy, txrc); + + info->flags |= mi->tx_flags; + sample_idx = minstrel_get_sample_rate(mp, mi); + if (sample_idx >= 0) { + minstrel_ht_set_rate(mp, mi, &ar[0], sample_idx, + txrc, true, false); + minstrel_ht_set_rate(mp, mi, &ar[1], mi->max_tp_rate, + txrc, false, true); + info->flags |= IEEE80211_TX_CTL_RATE_CTRL_PROBE; + } else { + minstrel_ht_set_rate(mp, mi, &ar[0], mi->max_tp_rate, + txrc, false, false); + minstrel_ht_set_rate(mp, mi, &ar[1], mi->max_tp_rate2, + txrc, false, true); + } + minstrel_ht_set_rate(mp, mi, &ar[2], mi->max_prob_rate, txrc, false, true); + + ar[3].count = 0; + ar[3].idx = -1; + + mi->total_packets++; + + /* wraparound */ + if (mi->total_packets == ~0) { + mi->total_packets = 0; + mi->sample_packets = 0; + } +} + +static void +minstrel_ht_update_caps(void *priv, struct ieee80211_supported_band *sband, + struct ieee80211_sta *sta, void *priv_sta, + enum nl80211_channel_type oper_chan_type) +{ + struct minstrel_priv *mp = priv; + struct minstrel_ht_sta_priv *msp = priv_sta; + struct minstrel_ht_sta *mi = &msp->ht; + struct ieee80211_mcs_info *mcs = &sta->ht_cap.mcs; + struct ieee80211_local *local = hw_to_local(mp->hw); + u16 sta_cap = sta->ht_cap.cap; + int ack_dur; + int stbc; + int i; + + /* fall back to the old minstrel for legacy stations */ + if (sta && !sta->ht_cap.ht_supported) { + msp->is_ht = false; + memset(&msp->legacy, 0, sizeof(msp->legacy)); + msp->legacy.r = msp->ratelist; + msp->legacy.sample_table = msp->sample_table; + return mac80211_minstrel.rate_init(priv, sband, sta, &msp->legacy); + } + + BUILD_BUG_ON(ARRAY_SIZE(minstrel_mcs_groups) != + MINSTREL_MAX_STREAMS * MINSTREL_STREAM_GROUPS); + + msp->is_ht = true; + memset(mi, 0, sizeof(*mi)); + mi->stats_update = jiffies; + + ack_dur = ieee80211_frame_duration(local, 10, 60, 1, 1); + mi->overhead = ieee80211_frame_duration(local, 0, 60, 1, 1) + ack_dur; + mi->overhead_rtscts = mi->overhead + 2 * ack_dur; + + mi->avg_ampdu_len = MINSTREL_FRAC(1, 1); + + /* When using MRR, sample more on the first attempt, without delay */ + if (mp->has_mrr) { + mi->sample_count = 16; + mi->sample_wait = 0; + } else { + mi->sample_count = 8; + mi->sample_wait = 8; + } + mi->sample_tries = 4; + + stbc = (sta_cap & IEEE80211_HT_CAP_RX_STBC) >> + IEEE80211_HT_CAP_RX_STBC_SHIFT; + mi->tx_flags |= stbc << IEEE80211_TX_CTL_STBC_SHIFT; + + if (sta_cap & IEEE80211_HT_CAP_LDPC_CODING) + mi->tx_flags |= IEEE80211_TX_CTL_LDPC; + + if (oper_chan_type != NL80211_CHAN_HT40MINUS && + oper_chan_type != NL80211_CHAN_HT40PLUS) + sta_cap &= ~IEEE80211_HT_CAP_SUP_WIDTH_20_40; + + for (i = 0; i < ARRAY_SIZE(mi->groups); i++) { + u16 req = 0; + + mi->groups[i].supported = 0; + if (minstrel_mcs_groups[i].flags & IEEE80211_TX_RC_SHORT_GI) { + if (minstrel_mcs_groups[i].flags & IEEE80211_TX_RC_40_MHZ_WIDTH) + req |= IEEE80211_HT_CAP_SGI_40; + else + req |= IEEE80211_HT_CAP_SGI_20; + } + + if (minstrel_mcs_groups[i].flags & IEEE80211_TX_RC_40_MHZ_WIDTH) + req |= IEEE80211_HT_CAP_SUP_WIDTH_20_40; + + if ((sta_cap & req) != req) + continue; + + mi->groups[i].supported = + mcs->rx_mask[minstrel_mcs_groups[i].streams - 1]; + } +} + +static void +minstrel_ht_rate_init(void *priv, struct ieee80211_supported_band *sband, + struct ieee80211_sta *sta, void *priv_sta) +{ + struct minstrel_priv *mp = priv; + + minstrel_ht_update_caps(priv, sband, sta, priv_sta, mp->hw->conf.channel_type); +} + +static void +minstrel_ht_rate_update(void *priv, struct ieee80211_supported_band *sband, + struct ieee80211_sta *sta, void *priv_sta, + u32 changed, enum nl80211_channel_type oper_chan_type) +{ + minstrel_ht_update_caps(priv, sband, sta, priv_sta, oper_chan_type); +} + +static void * +minstrel_ht_alloc_sta(void *priv, struct ieee80211_sta *sta, gfp_t gfp) +{ + struct ieee80211_supported_band *sband; + struct minstrel_ht_sta_priv *msp; + struct minstrel_priv *mp = priv; + struct ieee80211_hw *hw = mp->hw; + int max_rates = 0; + int i; + + for (i = 0; i < IEEE80211_NUM_BANDS; i++) { + sband = hw->wiphy->bands[i]; + if (sband && sband->n_bitrates > max_rates) + max_rates = sband->n_bitrates; + } + + msp = kzalloc(sizeof(struct minstrel_ht_sta), gfp); + if (!msp) + return NULL; + + msp->ratelist = kzalloc(sizeof(struct minstrel_rate) * max_rates, gfp); + if (!msp->ratelist) + goto error; + + msp->sample_table = kmalloc(SAMPLE_COLUMNS * max_rates, gfp); + if (!msp->sample_table) + goto error1; + + return msp; + +error1: + kfree(msp->sample_table); +error: + kfree(msp); + return NULL; +} + +static void +minstrel_ht_free_sta(void *priv, struct ieee80211_sta *sta, void *priv_sta) +{ + struct minstrel_ht_sta_priv *msp = priv_sta; + + kfree(msp->sample_table); + kfree(msp->ratelist); + kfree(msp); +} + +static void * +minstrel_ht_alloc(struct ieee80211_hw *hw, struct dentry *debugfsdir) +{ + return mac80211_minstrel.alloc(hw, debugfsdir); +} + +static void +minstrel_ht_free(void *priv) +{ + mac80211_minstrel.free(priv); +} + +static struct rate_control_ops mac80211_minstrel_ht = { + .name = "minstrel_ht", + .tx_status = minstrel_ht_tx_status, + .get_rate = minstrel_ht_get_rate, + .rate_init = minstrel_ht_rate_init, + .rate_update = minstrel_ht_rate_update, + .alloc_sta = minstrel_ht_alloc_sta, + .free_sta = minstrel_ht_free_sta, + .alloc = minstrel_ht_alloc, + .free = minstrel_ht_free, +#ifdef CONFIG_MAC80211_DEBUGFS + .add_sta_debugfs = minstrel_ht_add_sta_debugfs, + .remove_sta_debugfs = minstrel_ht_remove_sta_debugfs, +#endif +}; + + +static void +init_sample_table(void) +{ + int col, i, new_idx; + u8 rnd[MCS_GROUP_RATES]; + + memset(sample_table, 0xff, sizeof(sample_table)); + for (col = 0; col < SAMPLE_COLUMNS; col++) { + for (i = 0; i < MCS_GROUP_RATES; i++) { + get_random_bytes(rnd, sizeof(rnd)); + new_idx = (i + rnd[i]) % MCS_GROUP_RATES; + + while (sample_table[col][new_idx] != 0xff) + new_idx = (new_idx + 1) % MCS_GROUP_RATES; + + sample_table[col][new_idx] = i; + } + } +} + +int __init +rc80211_minstrel_ht_init(void) +{ + init_sample_table(); + return ieee80211_rate_control_register(&mac80211_minstrel_ht); +} + +void +rc80211_minstrel_ht_exit(void) +{ + ieee80211_rate_control_unregister(&mac80211_minstrel_ht); +} diff --git a/net/mac80211/rc80211_minstrel_ht.h b/net/mac80211/rc80211_minstrel_ht.h new file mode 100644 index 0000000..696c0fc --- /dev/null +++ b/net/mac80211/rc80211_minstrel_ht.h @@ -0,0 +1,128 @@ +/* + * Copyright (C) 2010 Felix Fietkau + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef __RC_MINSTREL_HT_H +#define __RC_MINSTREL_HT_H + +/* + * The number of streams can be changed to 2 to reduce code + * size and memory footprint. + */ +#define MINSTREL_MAX_STREAMS 3 +#define MINSTREL_STREAM_GROUPS 4 + +/* scaled fraction values */ +#define MINSTREL_SCALE 16 +#define MINSTREL_FRAC(val, div) (((val) << MINSTREL_SCALE) / div) +#define MINSTREL_TRUNC(val) ((val) >> MINSTREL_SCALE) + +#define MCS_GROUP_RATES 8 + +struct mcs_group { + u32 flags; + unsigned int streams; + unsigned int duration[MCS_GROUP_RATES]; +}; + +struct minstrel_rate_stats { + /* current / last sampling period attempts/success counters */ + unsigned int attempts, last_attempts; + unsigned int success, last_success; + + /* total attempts/success counters */ + u64 att_hist, succ_hist; + + /* current throughput */ + unsigned int cur_tp; + + /* packet delivery probabilities */ + unsigned int cur_prob, probability; + + /* maximum retry counts */ + unsigned int retry_count; + unsigned int retry_count_rtscts; + + bool retry_updated; + u8 sample_skipped; +}; + +struct minstrel_mcs_group_data { + u8 index; + u8 column; + + /* bitfield of supported MCS rates of this group */ + u8 supported; + + /* selected primary rates */ + unsigned int max_tp_rate; + unsigned int max_tp_rate2; + unsigned int max_prob_rate; + + /* MCS rate statistics */ + struct minstrel_rate_stats rates[MCS_GROUP_RATES]; +}; + +struct minstrel_ht_sta { + /* ampdu length (average, per sampling interval) */ + unsigned int ampdu_len; + unsigned int ampdu_packets; + + /* ampdu length (EWMA) */ + unsigned int avg_ampdu_len; + + /* best throughput rate */ + unsigned int max_tp_rate; + + /* second best throughput rate */ + unsigned int max_tp_rate2; + + /* best probability rate */ + unsigned int max_prob_rate; + + /* time of last status update */ + unsigned long stats_update; + + /* overhead time in usec for each frame */ + unsigned int overhead; + unsigned int overhead_rtscts; + + unsigned int total_packets; + unsigned int sample_packets; + + /* tx flags to add for frames for this sta */ + u32 tx_flags; + + u8 sample_wait; + u8 sample_tries; + u8 sample_count; + u8 sample_slow; + + /* current MCS group to be sampled */ + u8 sample_group; + + /* MCS rate group info and statistics */ + struct minstrel_mcs_group_data groups[MINSTREL_MAX_STREAMS * MINSTREL_STREAM_GROUPS]; +}; + +struct minstrel_ht_sta_priv { + union { + struct minstrel_ht_sta ht; + struct minstrel_sta_info legacy; + }; +#ifdef CONFIG_MAC80211_DEBUGFS + struct dentry *dbg_stats; +#endif + void *ratelist; + void *sample_table; + bool is_ht; +}; + +void minstrel_ht_add_sta_debugfs(void *priv, void *priv_sta, struct dentry *dir); +void minstrel_ht_remove_sta_debugfs(void *priv, void *priv_sta); + +#endif diff --git a/net/mac80211/rc80211_minstrel_ht_debugfs.c b/net/mac80211/rc80211_minstrel_ht_debugfs.c new file mode 100644 index 0000000..4fb3ccb --- /dev/null +++ b/net/mac80211/rc80211_minstrel_ht_debugfs.c @@ -0,0 +1,120 @@ +/* + * Copyright (C) 2010 Felix Fietkau + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include +#include +#include +#include +#include +#include +#include "rc80211_minstrel.h" +#include "rc80211_minstrel_ht.h" + +extern const struct mcs_group minstrel_mcs_groups[]; + +static int +minstrel_ht_stats_open(struct inode *inode, struct file *file) +{ + struct minstrel_ht_sta_priv *msp = inode->i_private; + struct minstrel_ht_sta *mi = &msp->ht; + struct minstrel_debugfs_info *ms; + unsigned int i, j, tp, prob, eprob; + char *p; + int ret; + + if (!msp->is_ht) { + inode->i_private = &msp->legacy; + ret = minstrel_stats_open(inode, file); + inode->i_private = msp; + return ret; + } + + ms = kmalloc(sizeof(*ms) + 8192, GFP_KERNEL); + if (!ms) + return -ENOMEM; + + file->private_data = ms; + p = ms->buf; + p += sprintf(p, "type rate throughput ewma prob this prob " + "this succ/attempt success attempts\n"); + for (i = 0; i < MINSTREL_MAX_STREAMS * MINSTREL_STREAM_GROUPS; i++) { + char htmode = '2'; + char gimode = 'L'; + + if (!mi->groups[i].supported) + continue; + + if (minstrel_mcs_groups[i].flags & IEEE80211_TX_RC_40_MHZ_WIDTH) + htmode = '4'; + if (minstrel_mcs_groups[i].flags & IEEE80211_TX_RC_SHORT_GI) + gimode = 'S'; + + for (j = 0; j < MCS_GROUP_RATES; j++) { + struct minstrel_rate_stats *mr = &mi->groups[i].rates[j]; + int idx = i * MCS_GROUP_RATES + j; + + if (!(mi->groups[i].supported & BIT(j))) + continue; + + p += sprintf(p, "HT%c0/%cGI ", htmode, gimode); + + *(p++) = (idx == mi->max_tp_rate) ? 'T' : ' '; + *(p++) = (idx == mi->max_tp_rate2) ? 't' : ' '; + *(p++) = (idx == mi->max_prob_rate) ? 'P' : ' '; + p += sprintf(p, "MCS%-2u", (minstrel_mcs_groups[i].streams - 1) * + MCS_GROUP_RATES + j); + + tp = mr->cur_tp / 10; + prob = MINSTREL_TRUNC(mr->cur_prob * 1000); + eprob = MINSTREL_TRUNC(mr->probability * 1000); + + p += sprintf(p, " %6u.%1u %6u.%1u %6u.%1u " + "%3u(%3u) %8llu %8llu\n", + tp / 10, tp % 10, + eprob / 10, eprob % 10, + prob / 10, prob % 10, + mr->last_success, + mr->last_attempts, + (unsigned long long)mr->succ_hist, + (unsigned long long)mr->att_hist); + } + } + p += sprintf(p, "\nTotal packet count:: ideal %d " + "lookaround %d\n", + max(0, (int) mi->total_packets - (int) mi->sample_packets), + mi->sample_packets); + p += sprintf(p, "Average A-MPDU length: %d.%d\n", + MINSTREL_TRUNC(mi->avg_ampdu_len), + MINSTREL_TRUNC(mi->avg_ampdu_len * 10) % 10); + ms->len = p - ms->buf; + + return 0; +} + +static const struct file_operations minstrel_ht_stat_fops = { + .owner = THIS_MODULE, + .open = minstrel_ht_stats_open, + .read = minstrel_stats_read, + .release = minstrel_stats_release, +}; + +void +minstrel_ht_add_sta_debugfs(void *priv, void *priv_sta, struct dentry *dir) +{ + struct minstrel_ht_sta_priv *msp = priv_sta; + + msp->dbg_stats = debugfs_create_file("rc_stats", S_IRUGO, dir, msp, + &minstrel_ht_stat_fops); +} + +void +minstrel_ht_remove_sta_debugfs(void *priv, void *priv_sta) +{ + struct minstrel_ht_sta_priv *msp = priv_sta; + + debugfs_remove(msp->dbg_stats); +} -- cgit v1.1 From aed8e1f9910b216ab3e14cb286c431c870f9b78f Mon Sep 17 00:00:00 2001 From: Helmut Schaa Date: Mon, 17 May 2010 17:30:59 +0200 Subject: cfg80211: don't refuse HT20 channels on devices that don't support HT40 Don't refuse HT20 channels on devices that don't support HT40. Signed-off-by: Helmut Schaa Signed-off-by: John W. Linville --- net/wireless/chan.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/wireless/chan.c b/net/wireless/chan.c index b01a6f6..d0c92dd 100644 --- a/net/wireless/chan.c +++ b/net/wireless/chan.c @@ -35,8 +35,9 @@ rdev_freq_to_chan(struct cfg80211_registered_device *rdev, if (!ht_cap->ht_supported) return NULL; - if (!(ht_cap->cap & IEEE80211_HT_CAP_SUP_WIDTH_20_40) || - ht_cap->cap & IEEE80211_HT_CAP_40MHZ_INTOLERANT) + if (channel_type != NL80211_CHAN_HT20 && + (!(ht_cap->cap & IEEE80211_HT_CAP_SUP_WIDTH_20_40) || + ht_cap->cap & IEEE80211_HT_CAP_40MHZ_INTOLERANT)) return NULL; } -- cgit v1.1 From 77c2061d10a408d0220c2b0e7faefe52d9c41008 Mon Sep 17 00:00:00 2001 From: Walter Goldens Date: Tue, 18 May 2010 04:44:54 -0700 Subject: wireless: fix several minor description typos Signed-off-by: Walter Goldens Signed-off-by: John W. Linville --- net/mac80211/mlme.c | 2 +- net/mac80211/status.c | 2 +- net/mac80211/work.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) (limited to 'net') diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 0839c4e..31e3386 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1763,7 +1763,7 @@ static void ieee80211_sta_work(struct work_struct *work) /* * ieee80211_queue_work() should have picked up most cases, - * here we'll pick the the rest. + * here we'll pick the rest. */ if (WARN(local->suspended, "STA MLME work scheduled while " "going to suspend\n")) diff --git a/net/mac80211/status.c b/net/mac80211/status.c index 94613af0..34da679 100644 --- a/net/mac80211/status.c +++ b/net/mac80211/status.c @@ -47,7 +47,7 @@ static void ieee80211_handle_filtered_frame(struct ieee80211_local *local, /* * This skb 'survived' a round-trip through the driver, and * hopefully the driver didn't mangle it too badly. However, - * we can definitely not rely on the the control information + * we can definitely not rely on the control information * being correct. Clear it so we don't get junk there, and * indicate that it needs new processing, but must not be * modified/encrypted again. diff --git a/net/mac80211/work.c b/net/mac80211/work.c index be3d4a69..4157717 100644 --- a/net/mac80211/work.c +++ b/net/mac80211/work.c @@ -840,7 +840,7 @@ static void ieee80211_work_work(struct work_struct *work) /* * ieee80211_queue_work() should have picked up most cases, - * here we'll pick the the rest. + * here we'll pick the rest. */ if (WARN(local->suspended, "work scheduled while going to suspend\n")) return; -- cgit v1.1 From 252aa631f88080920a7083ac5a5844ffc5463629 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 19 May 2010 12:17:12 +0200 Subject: cfg80211: make action channel type optional When sending action frames, we want to verify that we do that on the correct channel. However, checking the channel type in addition can get in the way, since the channel type could change on the fly during an association, and it's not useful to have the channel type anyway since it has no effect on the transmission. Therefore, make it optional to specify so that if wanted, it can still be checked, but is not required. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/cfg.c | 4 +++- net/mac80211/ieee80211_i.h | 1 + net/mac80211/mlme.c | 9 ++++++--- net/wireless/core.h | 1 + net/wireless/mlme.c | 3 ++- net/wireless/nl80211.c | 3 +++ 6 files changed, 16 insertions(+), 5 deletions(-) (limited to 'net') diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index c7000a6..f8c49c5 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1554,10 +1554,12 @@ static int ieee80211_cancel_remain_on_channel(struct wiphy *wiphy, static int ieee80211_action(struct wiphy *wiphy, struct net_device *dev, struct ieee80211_channel *chan, enum nl80211_channel_type channel_type, + bool channel_type_valid, const u8 *buf, size_t len, u64 *cookie) { return ieee80211_mgd_action(IEEE80211_DEV_TO_SUB_IF(dev), chan, - channel_type, buf, len, cookie); + channel_type, channel_type_valid, + buf, len, cookie); } struct cfg80211_ops mac80211_config_ops = { diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 1a9e2da..d4677ef 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -988,6 +988,7 @@ int ieee80211_mgd_disassoc(struct ieee80211_sub_if_data *sdata, int ieee80211_mgd_action(struct ieee80211_sub_if_data *sdata, struct ieee80211_channel *chan, enum nl80211_channel_type channel_type, + bool channel_type_valid, const u8 *buf, size_t len, u64 *cookie); ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 31e3386..29c3a75 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2308,6 +2308,7 @@ int ieee80211_mgd_disassoc(struct ieee80211_sub_if_data *sdata, int ieee80211_mgd_action(struct ieee80211_sub_if_data *sdata, struct ieee80211_channel *chan, enum nl80211_channel_type channel_type, + bool channel_type_valid, const u8 *buf, size_t len, u64 *cookie) { struct ieee80211_local *local = sdata->local; @@ -2315,9 +2316,11 @@ int ieee80211_mgd_action(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb; /* Check that we are on the requested channel for transmission */ - if ((chan != local->tmp_channel || - channel_type != local->tmp_channel_type) && - (chan != local->oper_channel || + if (chan != local->tmp_channel && + chan != local->oper_channel) + return -EBUSY; + if (channel_type_valid && + (channel_type != local->tmp_channel_type && channel_type != local->_oper_channel_type)) return -EBUSY; diff --git a/net/wireless/core.h b/net/wireless/core.h index ae930ac..63d57ae 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -339,6 +339,7 @@ int cfg80211_mlme_action(struct cfg80211_registered_device *rdev, struct net_device *dev, struct ieee80211_channel *chan, enum nl80211_channel_type channel_type, + bool channel_type_valid, const u8 *buf, size_t len, u64 *cookie); /* SME */ diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 48ead6f..f69ae19 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -827,6 +827,7 @@ int cfg80211_mlme_action(struct cfg80211_registered_device *rdev, struct net_device *dev, struct ieee80211_channel *chan, enum nl80211_channel_type channel_type, + bool channel_type_valid, const u8 *buf, size_t len, u64 *cookie) { struct wireless_dev *wdev = dev->ieee80211_ptr; @@ -855,7 +856,7 @@ int cfg80211_mlme_action(struct cfg80211_registered_device *rdev, /* Transmit the Action frame as requested by user space */ return rdev->ops->action(&rdev->wiphy, dev, chan, channel_type, - buf, len, cookie); + channel_type_valid, buf, len, cookie); } bool cfg80211_rx_action(struct net_device *dev, int freq, const u8 *buf, diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index db71150..90ab3c8 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -4681,6 +4681,7 @@ static int nl80211_action(struct sk_buff *skb, struct genl_info *info) struct net_device *dev; struct ieee80211_channel *chan; enum nl80211_channel_type channel_type = NL80211_CHAN_NO_HT; + bool channel_type_valid = false; u32 freq; int err; void *hdr; @@ -4722,6 +4723,7 @@ static int nl80211_action(struct sk_buff *skb, struct genl_info *info) err = -EINVAL; goto out; } + channel_type_valid = true; } freq = nla_get_u32(info->attrs[NL80211_ATTR_WIPHY_FREQ]); @@ -4745,6 +4747,7 @@ static int nl80211_action(struct sk_buff *skb, struct genl_info *info) goto free_msg; } err = cfg80211_mlme_action(rdev, dev, chan, channel_type, + channel_type_valid, nla_data(info->attrs[NL80211_ATTR_FRAME]), nla_len(info->attrs[NL80211_ATTR_FRAME]), &cookie); -- cgit v1.1 From b5f7e7554753e2cc3ef3bef0271fdb32027df2ba Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 2 Jun 2010 12:05:27 +0000 Subject: ipv4: add LINUX_MIB_IPRPFILTER snmp counter Christoph Lameter mentioned that packets could be dropped in input path because of rp_filter settings, without any SNMP counter being incremented. System administrator can have a hard time to track the problem. This patch introduces a new counter, LINUX_MIB_IPRPFILTER, incremented each time we drop a packet because Reverse Path Filter triggers. (We receive an IPv4 datagram on a given interface, and find the route to send an answer would use another interface) netstat -s | grep IPReversePathFilter IPReversePathFilter: 21714 Reported-by: Christoph Lameter Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/fib_frontend.c | 6 ++++-- net/ipv4/ip_input.c | 3 +++ net/ipv4/proc.c | 1 + net/ipv4/route.c | 31 ++++++++++++++++++------------- 4 files changed, 26 insertions(+), 15 deletions(-) (limited to 'net') diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c index 4f0ed45..e830f7a 100644 --- a/net/ipv4/fib_frontend.c +++ b/net/ipv4/fib_frontend.c @@ -284,7 +284,7 @@ int fib_validate_source(__be32 src, __be32 dst, u8 tos, int oif, if (no_addr) goto last_resort; if (rpf == 1) - goto e_inval; + goto e_rpf; fl.oif = dev->ifindex; ret = 0; @@ -299,7 +299,7 @@ int fib_validate_source(__be32 src, __be32 dst, u8 tos, int oif, last_resort: if (rpf) - goto e_inval; + goto e_rpf; *spec_dst = inet_select_addr(dev, 0, RT_SCOPE_UNIVERSE); *itag = 0; return 0; @@ -308,6 +308,8 @@ e_inval_res: fib_res_put(&res); e_inval: return -EINVAL; +e_rpf: + return -EXDEV; } static inline __be32 sk_extract_addr(struct sockaddr *addr) diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c index d930dc5..d52c9da 100644 --- a/net/ipv4/ip_input.c +++ b/net/ipv4/ip_input.c @@ -340,6 +340,9 @@ static int ip_rcv_finish(struct sk_buff *skb) else if (err == -ENETUNREACH) IP_INC_STATS_BH(dev_net(skb->dev), IPSTATS_MIB_INNOROUTES); + else if (err == -EXDEV) + NET_INC_STATS_BH(dev_net(skb->dev), + LINUX_MIB_IPRPFILTER); goto drop; } } diff --git a/net/ipv4/proc.c b/net/ipv4/proc.c index 3dc9914..e320ca6 100644 --- a/net/ipv4/proc.c +++ b/net/ipv4/proc.c @@ -252,6 +252,7 @@ static const struct snmp_mib snmp4_net_list[] = { SNMP_MIB_ITEM("TCPBacklogDrop", LINUX_MIB_TCPBACKLOGDROP), SNMP_MIB_ITEM("TCPMinTTLDrop", LINUX_MIB_TCPMINTTLDROP), SNMP_MIB_ITEM("TCPDeferAcceptDrop", LINUX_MIB_TCPDEFERACCEPTDROP), + SNMP_MIB_ITEM("IPReversePathFilter", LINUX_MIB_IPRPFILTER), SNMP_MIB_SENTINEL }; diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 8495bce..d377b45 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -1851,6 +1851,7 @@ static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr, __be32 spec_dst; struct in_device *in_dev = in_dev_get(dev); u32 itag = 0; + int err; /* Primary sanity checks. */ @@ -1865,10 +1866,12 @@ static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr, if (!ipv4_is_local_multicast(daddr)) goto e_inval; spec_dst = inet_select_addr(dev, 0, RT_SCOPE_LINK); - } else if (fib_validate_source(saddr, 0, tos, 0, - dev, &spec_dst, &itag, 0) < 0) - goto e_inval; - + } else { + err = fib_validate_source(saddr, 0, tos, 0, dev, &spec_dst, + &itag, 0); + if (err < 0) + goto e_err; + } rth = dst_alloc(&ipv4_dst_ops); if (!rth) goto e_nobufs; @@ -1920,8 +1923,10 @@ e_nobufs: return -ENOBUFS; e_inval: + err = -EINVAL; +e_err: in_dev_put(in_dev); - return -EINVAL; + return err; } @@ -1985,7 +1990,6 @@ static int __mkroute_input(struct sk_buff *skb, ip_handle_martian_source(in_dev->dev, in_dev, skb, daddr, saddr); - err = -EINVAL; goto cleanup; } @@ -2157,13 +2161,12 @@ static int ip_route_input_slow(struct sk_buff *skb, __be32 daddr, __be32 saddr, goto brd_input; if (res.type == RTN_LOCAL) { - int result; - result = fib_validate_source(saddr, daddr, tos, + err = fib_validate_source(saddr, daddr, tos, net->loopback_dev->ifindex, dev, &spec_dst, &itag, skb->mark); - if (result < 0) - goto martian_source; - if (result) + if (err < 0) + goto martian_source_keep_err; + if (err) flags |= RTCF_DIRECTSRC; spec_dst = daddr; goto local_input; @@ -2191,7 +2194,7 @@ brd_input: err = fib_validate_source(saddr, 0, tos, 0, dev, &spec_dst, &itag, skb->mark); if (err < 0) - goto martian_source; + goto martian_source_keep_err; if (err) flags |= RTCF_DIRECTSRC; } @@ -2272,8 +2275,10 @@ e_nobufs: goto done; martian_source: + err = -EINVAL; +martian_source_keep_err: ip_handle_martian_source(dev, in_dev, skb, daddr, saddr); - goto e_inval; + goto done; } int ip_route_input_common(struct sk_buff *skb, __be32 daddr, __be32 saddr, -- cgit v1.1 From 96d362202bfc0e5da78ee59b1645296fbca515f4 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 2 Jun 2010 19:21:31 +0000 Subject: ipv4: RCU conversion of ip_route_input_slow/ip_route_input_mc Avoid two atomic ops on struct in_device refcount per incoming packet, if slow path taken, (or route cache disabled) Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/route.c | 35 +++++++++++++++++------------------ 1 file changed, 17 insertions(+), 18 deletions(-) (limited to 'net') diff --git a/net/ipv4/route.c b/net/ipv4/route.c index d377b45..1cfe0d1 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -1843,13 +1843,14 @@ static void rt_set_nexthop(struct rtable *rt, struct fib_result *res, u32 itag) rt->rt_type = res->type; } +/* called in rcu_read_lock() section */ static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr, u8 tos, struct net_device *dev, int our) { - unsigned hash; + unsigned int hash; struct rtable *rth; __be32 spec_dst; - struct in_device *in_dev = in_dev_get(dev); + struct in_device *in_dev = __in_dev_get_rcu(dev); u32 itag = 0; int err; @@ -1914,18 +1915,14 @@ static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr, #endif RT_CACHE_STAT_INC(in_slow_mc); - in_dev_put(in_dev); hash = rt_hash(daddr, saddr, dev->ifindex, rt_genid(dev_net(dev))); return rt_intern_hash(hash, rth, NULL, skb, dev->ifindex); e_nobufs: - in_dev_put(in_dev); return -ENOBUFS; - e_inval: - err = -EINVAL; + return -EINVAL; e_err: - in_dev_put(in_dev); return err; } @@ -2101,7 +2098,7 @@ static int ip_route_input_slow(struct sk_buff *skb, __be32 daddr, __be32 saddr, u8 tos, struct net_device *dev) { struct fib_result res; - struct in_device *in_dev = in_dev_get(dev); + struct in_device *in_dev = __in_dev_get_rcu(dev); struct flowi fl = { .nl_u = { .ip4_u = { .daddr = daddr, .saddr = saddr, @@ -2179,7 +2176,6 @@ static int ip_route_input_slow(struct sk_buff *skb, __be32 daddr, __be32 saddr, err = ip_mkroute_input(skb, &res, &fl, in_dev, daddr, saddr, tos); done: - in_dev_put(in_dev); if (free_res) fib_res_put(&res); out: return err; @@ -2288,16 +2284,18 @@ int ip_route_input_common(struct sk_buff *skb, __be32 daddr, __be32 saddr, unsigned hash; int iif = dev->ifindex; struct net *net; + int res; net = dev_net(dev); + rcu_read_lock(); + if (!rt_caching(net)) goto skip_cache; tos &= IPTOS_RT_MASK; hash = rt_hash(daddr, saddr, iif, rt_genid(net)); - rcu_read_lock(); for (rth = rcu_dereference(rt_hash_table[hash].chain); rth; rth = rcu_dereference(rth->u.dst.rt_next)) { if ((((__force u32)rth->fl.fl4_dst ^ (__force u32)daddr) | @@ -2321,7 +2319,6 @@ int ip_route_input_common(struct sk_buff *skb, __be32 daddr, __be32 saddr, } RT_CACHE_STAT_INC(in_hlist_search); } - rcu_read_unlock(); skip_cache: /* Multicast recognition logic is moved from route cache to here. @@ -2336,12 +2333,11 @@ skip_cache: route cache entry is created eventually. */ if (ipv4_is_multicast(daddr)) { - struct in_device *in_dev; + struct in_device *in_dev = __in_dev_get_rcu(dev); - rcu_read_lock(); - if ((in_dev = __in_dev_get_rcu(dev)) != NULL) { + if (in_dev) { int our = ip_check_mc(in_dev, daddr, saddr, - ip_hdr(skb)->protocol); + ip_hdr(skb)->protocol); if (our #ifdef CONFIG_IP_MROUTE || @@ -2349,15 +2345,18 @@ skip_cache: IN_DEV_MFORWARD(in_dev)) #endif ) { + int res = ip_route_input_mc(skb, daddr, saddr, + tos, dev, our); rcu_read_unlock(); - return ip_route_input_mc(skb, daddr, saddr, - tos, dev, our); + return res; } } rcu_read_unlock(); return -EINVAL; } - return ip_route_input_slow(skb, daddr, saddr, tos, dev); + res = ip_route_input_slow(skb, daddr, saddr, tos, dev); + rcu_read_unlock(); + return res; } EXPORT_SYMBOL(ip_route_input_common); -- cgit v1.1 From bc10502dba37d3b210efd9f3867212298f13b78e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 3 Jun 2010 03:21:52 -0700 Subject: net: use __packed annotation cleanup patch. Use new __packed annotation in net/ and include/ (except netfilter) Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/bluetooth/bnep/bnep.h | 8 ++++---- net/compat.c | 6 +++--- net/iucv/iucv.c | 14 +++++++------- net/mac80211/cfg.c | 2 +- net/mac80211/ieee80211_i.h | 2 +- net/mac80211/rx.c | 2 +- net/sctp/sm_make_chunk.c | 2 +- 7 files changed, 18 insertions(+), 18 deletions(-) (limited to 'net') diff --git a/net/bluetooth/bnep/bnep.h b/net/bluetooth/bnep/bnep.h index 0d9e506..70672544 100644 --- a/net/bluetooth/bnep/bnep.h +++ b/net/bluetooth/bnep/bnep.h @@ -86,26 +86,26 @@ struct bnep_setup_conn_req { __u8 ctrl; __u8 uuid_size; __u8 service[0]; -} __attribute__((packed)); +} __packed; struct bnep_set_filter_req { __u8 type; __u8 ctrl; __be16 len; __u8 list[0]; -} __attribute__((packed)); +} __packed; struct bnep_control_rsp { __u8 type; __u8 ctrl; __be16 resp; -} __attribute__((packed)); +} __packed; struct bnep_ext_hdr { __u8 type; __u8 len; __u8 data[0]; -} __attribute__((packed)); +} __packed; /* BNEP ioctl defines */ #define BNEPCONNADD _IOW('B', 200, int) diff --git a/net/compat.c b/net/compat.c index ec24d9e..1cf7590 100644 --- a/net/compat.c +++ b/net/compat.c @@ -531,7 +531,7 @@ struct compat_group_req { __u32 gr_interface; struct __kernel_sockaddr_storage gr_group __attribute__ ((aligned(4))); -} __attribute__ ((packed)); +} __packed; struct compat_group_source_req { __u32 gsr_interface; @@ -539,7 +539,7 @@ struct compat_group_source_req { __attribute__ ((aligned(4))); struct __kernel_sockaddr_storage gsr_source __attribute__ ((aligned(4))); -} __attribute__ ((packed)); +} __packed; struct compat_group_filter { __u32 gf_interface; @@ -549,7 +549,7 @@ struct compat_group_filter { __u32 gf_numsrc; struct __kernel_sockaddr_storage gf_slist[1] __attribute__ ((aligned(4))); -} __attribute__ ((packed)); +} __packed; #define __COMPAT_GF0_SIZE (sizeof(struct compat_group_filter) - \ sizeof(struct __kernel_sockaddr_storage)) diff --git a/net/iucv/iucv.c b/net/iucv/iucv.c index f28ad2cc..499c045 100644 --- a/net/iucv/iucv.c +++ b/net/iucv/iucv.c @@ -1463,7 +1463,7 @@ struct iucv_path_pending { u32 res3; u8 ippollfg; u8 res4[3]; -} __attribute__ ((packed)); +} __packed; static void iucv_path_pending(struct iucv_irq_data *data) { @@ -1524,7 +1524,7 @@ struct iucv_path_complete { u32 res3; u8 ippollfg; u8 res4[3]; -} __attribute__ ((packed)); +} __packed; static void iucv_path_complete(struct iucv_irq_data *data) { @@ -1554,7 +1554,7 @@ struct iucv_path_severed { u32 res4; u8 ippollfg; u8 res5[3]; -} __attribute__ ((packed)); +} __packed; static void iucv_path_severed(struct iucv_irq_data *data) { @@ -1590,7 +1590,7 @@ struct iucv_path_quiesced { u32 res4; u8 ippollfg; u8 res5[3]; -} __attribute__ ((packed)); +} __packed; static void iucv_path_quiesced(struct iucv_irq_data *data) { @@ -1618,7 +1618,7 @@ struct iucv_path_resumed { u32 res4; u8 ippollfg; u8 res5[3]; -} __attribute__ ((packed)); +} __packed; static void iucv_path_resumed(struct iucv_irq_data *data) { @@ -1649,7 +1649,7 @@ struct iucv_message_complete { u32 ipbfln2f; u8 ippollfg; u8 res2[3]; -} __attribute__ ((packed)); +} __packed; static void iucv_message_complete(struct iucv_irq_data *data) { @@ -1694,7 +1694,7 @@ struct iucv_message_pending { u32 ipbfln2f; u8 ippollfg; u8 res2[3]; -} __attribute__ ((packed)); +} __packed; static void iucv_message_pending(struct iucv_irq_data *data) { diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index c7000a6..a2ed0f7b 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -600,7 +600,7 @@ struct iapp_layer2_update { u8 ssap; /* 0 */ u8 control; u8 xid_info[3]; -} __attribute__ ((packed)); +} __packed; static void ieee80211_send_layer2_update(struct sta_info *sta) { diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 1a9e2da..ec3e5c3 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -1084,7 +1084,7 @@ struct ieee80211_tx_status_rtap_hdr { u8 padding_for_rate; __le16 tx_flags; u8 data_retries; -} __attribute__ ((packed)); +} __packed; /* HT */ diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 6e2a7bc..2d9a2ee 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -2139,7 +2139,7 @@ static void ieee80211_rx_cooked_monitor(struct ieee80211_rx_data *rx, u8 rate_or_pad; __le16 chan_freq; __le16 chan_flags; - } __attribute__ ((packed)) *rthdr; + } __packed *rthdr; struct sk_buff *skb = rx->skb, *skb2; struct net_device *prev_dev = NULL; struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c index bd2a50b..246f929 100644 --- a/net/sctp/sm_make_chunk.c +++ b/net/sctp/sm_make_chunk.c @@ -1817,7 +1817,7 @@ malformed: struct __sctp_missing { __be32 num_missing; __be16 type; -} __attribute__((packed)); +} __packed; /* * Report a missing mandatory parameter. -- cgit v1.1 From 3a4d4aa2d38e5305b5e93dffdc9dd2f975129328 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 26 May 2010 16:41:40 +0200 Subject: mac80211: remove bogus mod_timer() call The IBSS code has a bogus mod_timer(..., 0) call, we shouldn't ever pass a constant value to the function since any constant value could be in the future or the past. However, invoking the timer here is not necessary at all, since we just finished scanning and just need to have the IBSS code run again from the workqueue later, so factor out the work starting and use that instead. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/ibss.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) (limited to 'net') diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index b2cc1fd..d7a96ce 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -798,6 +798,15 @@ static void ieee80211_ibss_work(struct work_struct *work) } } +static void ieee80211_queue_ibss_work(struct ieee80211_sub_if_data *sdata) +{ + struct ieee80211_if_ibss *ifibss = &sdata->u.ibss; + struct ieee80211_local *local = sdata->local; + + set_bit(IEEE80211_IBSS_REQ_RUN, &ifibss->request); + ieee80211_queue_work(&local->hw, &ifibss->work); +} + static void ieee80211_ibss_timer(unsigned long data) { struct ieee80211_sub_if_data *sdata = @@ -810,8 +819,7 @@ static void ieee80211_ibss_timer(unsigned long data) return; } - set_bit(IEEE80211_IBSS_REQ_RUN, &ifibss->request); - ieee80211_queue_work(&local->hw, &ifibss->work); + ieee80211_queue_ibss_work(sdata); } #ifdef CONFIG_PM @@ -859,7 +867,7 @@ void ieee80211_ibss_notify_scan_completed(struct ieee80211_local *local) if (!sdata->u.ibss.ssid_len) continue; sdata->u.ibss.last_scan_completed = jiffies; - mod_timer(&sdata->u.ibss.timer, 0); + ieee80211_queue_ibss_work(sdata); } mutex_unlock(&local->iflist_mtx); } -- cgit v1.1 From 2b2c009ecf71f4c66ff8420b63dddbc9737e04e3 Mon Sep 17 00:00:00 2001 From: Juuso Oikarinen Date: Thu, 27 May 2010 15:32:13 +0300 Subject: mac80211: Add support for hardware ARP query filtering Some hardware allow extended filtering of ARP frames not intended for the host. To perform such filtering, the hardware needs to know the current IP address(es) of the host, bound to its interface. Add support for ARP filtering to mac80211 by adding a new op to the driver interface, allowing to configure the current IP addresses. This op is called upon association with the currently configured address(es), and when associated whenever the IP address(es) change. This patch adds configuration of IPv4 addresses only, as IPv6 addresses don't need ARP filtering. Signed-off-by: Juuso Oikarinen Reviewed-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/driver-ops.h | 17 ++++++++++++ net/mac80211/driver-trace.h | 25 ++++++++++++++++++ net/mac80211/ieee80211_i.h | 2 ++ net/mac80211/main.c | 63 ++++++++++++++++++++++++++++++++++++++++++++- net/mac80211/mlme.c | 11 +++++++- 5 files changed, 116 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/mac80211/driver-ops.h b/net/mac80211/driver-ops.h index 4f22713..978850e 100644 --- a/net/mac80211/driver-ops.h +++ b/net/mac80211/driver-ops.h @@ -83,6 +83,23 @@ static inline void drv_bss_info_changed(struct ieee80211_local *local, trace_drv_bss_info_changed(local, sdata, info, changed); } +struct in_ifaddr; +static inline int drv_configure_arp_filter(struct ieee80211_local *local, + struct ieee80211_vif *vif, + struct in_ifaddr *ifa_list) +{ + int ret = 0; + + might_sleep(); + + if (local->ops->configure_arp_filter) + ret = local->ops->configure_arp_filter(&local->hw, vif, + ifa_list); + + trace_drv_configure_arp_filter(local, vif_to_sdata(vif), ifa_list, ret); + return ret; +} + static inline u64 drv_prepare_multicast(struct ieee80211_local *local, struct netdev_hw_addr_list *mc_list) { diff --git a/net/mac80211/driver-trace.h b/net/mac80211/driver-trace.h index 6a9b234..577460d 100644 --- a/net/mac80211/driver-trace.h +++ b/net/mac80211/driver-trace.h @@ -219,6 +219,31 @@ TRACE_EVENT(drv_bss_info_changed, ) ); +TRACE_EVENT(drv_configure_arp_filter, + TP_PROTO(struct ieee80211_local *local, + struct ieee80211_sub_if_data *sdata, + struct in_ifaddr *ifa_list, int ret), + + TP_ARGS(local, sdata, ifa_list, ret), + + TP_STRUCT__entry( + LOCAL_ENTRY + VIF_ENTRY + __field(int, ret) + ), + + TP_fast_assign( + LOCAL_ASSIGN; + VIF_ASSIGN; + __entry->ret = ret; + ), + + TP_printk( + VIF_PR_FMT LOCAL_PR_FMT " ret:%d", + VIF_PR_ARG, LOCAL_PR_ARG, __entry->ret + ) +); + TRACE_EVENT(drv_prepare_multicast, TP_PROTO(struct ieee80211_local *local, int mc_count, u64 ret), diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index d4677ef..47d6753 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -851,6 +851,7 @@ struct ieee80211_local { struct work_struct dynamic_ps_disable_work; struct timer_list dynamic_ps_timer; struct notifier_block network_latency_notifier; + struct notifier_block ifa_notifier; int user_power_level; /* in dBm */ int power_constr_level; /* in dBm */ @@ -997,6 +998,7 @@ void ieee80211_send_pspoll(struct ieee80211_local *local, void ieee80211_recalc_ps(struct ieee80211_local *local, s32 latency); int ieee80211_max_network_latency(struct notifier_block *nb, unsigned long data, void *dummy); +int ieee80211_set_arp_filter(struct ieee80211_sub_if_data *sdata); void ieee80211_sta_process_chanswitch(struct ieee80211_sub_if_data *sdata, struct ieee80211_channel_sw_ie *sw_elem, struct ieee80211_bss *bss, diff --git a/net/mac80211/main.c b/net/mac80211/main.c index c8548e61..4051b23 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -329,6 +329,58 @@ static void ieee80211_recalc_smps_work(struct work_struct *work) mutex_unlock(&local->iflist_mtx); } +int ieee80211_set_arp_filter(struct ieee80211_sub_if_data *sdata) +{ + struct in_device *idev; + int ret = 0; + + BUG_ON(!sdata); + ASSERT_RTNL(); + + idev = sdata->dev->ip_ptr; + if (!idev) + return 0; + + ret = drv_configure_arp_filter(sdata->local, &sdata->vif, + idev->ifa_list); + return ret; +} + +static int ieee80211_ifa_changed(struct notifier_block *nb, + unsigned long data, void *arg) +{ + struct in_ifaddr *ifa = arg; + struct ieee80211_local *local = + container_of(nb, struct ieee80211_local, + ifa_notifier); + struct net_device *ndev = ifa->ifa_dev->dev; + struct wireless_dev *wdev = ndev->ieee80211_ptr; + struct ieee80211_sub_if_data *sdata; + struct ieee80211_if_managed *ifmgd; + + /* Make sure it's our interface that got changed */ + if (!wdev) + return NOTIFY_DONE; + + if (wdev->wiphy != local->hw.wiphy) + return NOTIFY_DONE; + + /* We are concerned about IP addresses only when associated */ + sdata = IEEE80211_DEV_TO_SUB_IF(ndev); + + /* ARP filtering is only supported in managed mode */ + if (sdata->vif.type != NL80211_IFTYPE_STATION) + return NOTIFY_DONE; + + ifmgd = &sdata->u.mgd; + mutex_lock(&ifmgd->mtx); + if (ifmgd->associated) + ieee80211_set_arp_filter(sdata); + mutex_unlock(&ifmgd->mtx); + + return NOTIFY_DONE; +} + struct ieee80211_hw *ieee80211_alloc_hw(size_t priv_data_len, const struct ieee80211_ops *ops) { @@ -612,14 +664,22 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) ieee80211_max_network_latency; result = pm_qos_add_notifier(PM_QOS_NETWORK_LATENCY, &local->network_latency_notifier); - if (result) { rtnl_lock(); goto fail_pm_qos; } + local->ifa_notifier.notifier_call = ieee80211_ifa_changed; + result = register_inetaddr_notifier(&local->ifa_notifier); + if (result) + goto fail_ifa; + return 0; + fail_ifa: + pm_qos_remove_notifier(PM_QOS_NETWORK_LATENCY, + &local->network_latency_notifier); + rtnl_lock(); fail_pm_qos: ieee80211_led_exit(local); ieee80211_remove_interfaces(local); @@ -647,6 +707,7 @@ void ieee80211_unregister_hw(struct ieee80211_hw *hw) pm_qos_remove_notifier(PM_QOS_NETWORK_LATENCY, &local->network_latency_notifier); + unregister_inetaddr_notifier(&local->ifa_notifier); rtnl_lock(); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 29c3a75..7e72013 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2078,8 +2078,17 @@ static enum work_done_result ieee80211_assoc_done(struct ieee80211_work *wk, cfg80211_send_assoc_timeout(wk->sdata->dev, wk->filter_ta); return WORK_DONE_DESTROY; + } else { + mutex_unlock(&wk->sdata->u.mgd.mtx); + + /* + * configure ARP filter IP addresses to the driver, + * intentionally outside the mgd mutex. + */ + rtnl_lock(); + ieee80211_set_arp_filter(wk->sdata); + rtnl_unlock(); } - mutex_unlock(&wk->sdata->u.mgd.mtx); } cfg80211_send_rx_assoc(wk->sdata->dev, skb->data, skb->len); -- cgit v1.1 From 6a8579d0e62c0eac428184ce45e86bc46677724a Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 27 May 2010 14:41:07 +0200 Subject: mac80211: clean up ieee80211_stop_tx_ba_session There's no sense in letting anything but internal mac80211 functions set the initiator to anything but WLAN_BACK_INITIATOR, since WLAN_BACK_RECIPIENT is only valid when we have received a frame from the peer, which we react to directly in mac80211. The debugfs code I recently added got this wrong as well. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 7 +++---- net/mac80211/debugfs_sta.c | 3 +-- net/mac80211/driver-trace.h | 10 ++++------ 3 files changed, 8 insertions(+), 12 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index c163d0a..feb15c4 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -538,14 +538,13 @@ int __ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, return ret; } -int ieee80211_stop_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid, - enum ieee80211_back_parties initiator) +int ieee80211_stop_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) { struct sta_info *sta = container_of(pubsta, struct sta_info, sta); struct ieee80211_sub_if_data *sdata = sta->sdata; struct ieee80211_local *local = sdata->local; - trace_api_stop_tx_ba_session(pubsta, tid, initiator); + trace_api_stop_tx_ba_session(pubsta, tid); if (!local->ops->ampdu_action) return -EINVAL; @@ -553,7 +552,7 @@ int ieee80211_stop_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid, if (tid >= STA_TID_NUM) return -EINVAL; - return __ieee80211_stop_tx_ba_session(sta, tid, initiator); + return __ieee80211_stop_tx_ba_session(sta, tid, WLAN_BACK_INITIATOR); } EXPORT_SYMBOL(ieee80211_stop_tx_ba_session); diff --git a/net/mac80211/debugfs_sta.c b/net/mac80211/debugfs_sta.c index e763f15..9f14061 100644 --- a/net/mac80211/debugfs_sta.c +++ b/net/mac80211/debugfs_sta.c @@ -210,8 +210,7 @@ static ssize_t sta_agg_status_write(struct file *file, const char __user *userbu if (start) ret = ieee80211_start_tx_ba_session(&sta->sta, tid); else - ret = ieee80211_stop_tx_ba_session(&sta->sta, tid, - WLAN_BACK_RECIPIENT); + ret = ieee80211_stop_tx_ba_session(&sta->sta, tid); } else { __ieee80211_stop_rx_ba_session(sta, tid, WLAN_BACK_RECIPIENT, 3); ret = 0; diff --git a/net/mac80211/driver-trace.h b/net/mac80211/driver-trace.h index 577460d..6b906301 100644 --- a/net/mac80211/driver-trace.h +++ b/net/mac80211/driver-trace.h @@ -876,25 +876,23 @@ TRACE_EVENT(api_start_tx_ba_cb, ); TRACE_EVENT(api_stop_tx_ba_session, - TP_PROTO(struct ieee80211_sta *sta, u16 tid, u16 initiator), + TP_PROTO(struct ieee80211_sta *sta, u16 tid), - TP_ARGS(sta, tid, initiator), + TP_ARGS(sta, tid), TP_STRUCT__entry( STA_ENTRY __field(u16, tid) - __field(u16, initiator) ), TP_fast_assign( STA_ASSIGN; __entry->tid = tid; - __entry->initiator = initiator; ), TP_printk( - STA_PR_FMT " tid:%d initiator:%d", - STA_PR_ARG, __entry->tid, __entry->initiator + STA_PR_FMT " tid:%d", + STA_PR_ARG, __entry->tid ) ); -- cgit v1.1 From 761ab470364b550c9b1a5e1e31be51d415aaf42b Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 28 May 2010 14:24:19 +0200 Subject: mac80211: move WEP weak IV check I suspect the compiler will do this optimisation anyway, but it seems cleaner to move this into the WEP switch case. Also make rx_h_decrypt use a local variable for the frame_control so that we don't need to reload the hdr variable for this after linearizing. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/rx.c | 31 ++++++++++++++++--------------- 1 file changed, 16 insertions(+), 15 deletions(-) (limited to 'net') diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 6e2a7bc..0ade267 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -825,6 +825,7 @@ ieee80211_rx_h_decrypt(struct ieee80211_rx_data *rx) ieee80211_rx_result result = RX_DROP_UNUSABLE; struct ieee80211_key *stakey = NULL; int mmie_keyidx = -1; + __le16 fc; /* * Key selection 101 @@ -866,13 +867,15 @@ ieee80211_rx_h_decrypt(struct ieee80211_rx_data *rx) if (rx->sta) stakey = rcu_dereference(rx->sta->key); - if (!ieee80211_has_protected(hdr->frame_control)) + fc = hdr->frame_control; + + if (!ieee80211_has_protected(fc)) mmie_keyidx = ieee80211_get_mmie_keyidx(rx->skb); if (!is_multicast_ether_addr(hdr->addr1) && stakey) { rx->key = stakey; /* Skip decryption if the frame is not protected. */ - if (!ieee80211_has_protected(hdr->frame_control)) + if (!ieee80211_has_protected(fc)) return RX_CONTINUE; } else if (mmie_keyidx >= 0) { /* Broadcast/multicast robust management frame / BIP */ @@ -884,7 +887,7 @@ ieee80211_rx_h_decrypt(struct ieee80211_rx_data *rx) mmie_keyidx >= NUM_DEFAULT_KEYS + NUM_DEFAULT_MGMT_KEYS) return RX_DROP_MONITOR; /* unexpected BIP keyidx */ rx->key = rcu_dereference(rx->sdata->keys[mmie_keyidx]); - } else if (!ieee80211_has_protected(hdr->frame_control)) { + } else if (!ieee80211_has_protected(fc)) { /* * The frame was not protected, so skip decryption. However, we * need to set rx->key if there is a key that could have been @@ -892,7 +895,7 @@ ieee80211_rx_h_decrypt(struct ieee80211_rx_data *rx) * have been expected. */ struct ieee80211_key *key = NULL; - if (ieee80211_is_mgmt(hdr->frame_control) && + if (ieee80211_is_mgmt(fc) && is_multicast_ether_addr(hdr->addr1) && (key = rcu_dereference(rx->sdata->default_mgmt_key))) rx->key = key; @@ -914,7 +917,7 @@ ieee80211_rx_h_decrypt(struct ieee80211_rx_data *rx) (status->flag & RX_FLAG_IV_STRIPPED)) return RX_CONTINUE; - hdrlen = ieee80211_hdrlen(hdr->frame_control); + hdrlen = ieee80211_hdrlen(fc); if (rx->skb->len < 8 + hdrlen) return RX_DROP_UNUSABLE; /* TODO: count this? */ @@ -947,19 +950,17 @@ ieee80211_rx_h_decrypt(struct ieee80211_rx_data *rx) if (skb_linearize(rx->skb)) return RX_DROP_UNUSABLE; - - hdr = (struct ieee80211_hdr *)rx->skb->data; - - /* Check for weak IVs if possible */ - if (rx->sta && rx->key->conf.alg == ALG_WEP && - ieee80211_is_data(hdr->frame_control) && - (!(status->flag & RX_FLAG_IV_STRIPPED) || - !(status->flag & RX_FLAG_DECRYPTED)) && - ieee80211_wep_is_weak_iv(rx->skb, rx->key)) - rx->sta->wep_weak_iv_count++; + /* the hdr variable is invalid now! */ switch (rx->key->conf.alg) { case ALG_WEP: + /* Check for weak IVs if possible */ + if (rx->sta && ieee80211_is_data(fc) && + (!(status->flag & RX_FLAG_IV_STRIPPED) || + !(status->flag & RX_FLAG_DECRYPTED)) && + ieee80211_wep_is_weak_iv(rx->skb, rx->key)) + rx->sta->wep_weak_iv_count++; + result = ieee80211_crypto_wep_decrypt(rx); break; case ALG_TKIP: -- cgit v1.1 From e0961f112cd88176acc6d1af6ca6352f85cdf993 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 28 May 2010 14:48:52 +0200 Subject: mac80211: remove useless work starting Ever since we use only cfg80211 for configuration, there is no configuration that could be pending at this point, cfg80211 will have the configuration that is pending and apply it afterwards. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/iface.c | 9 --------- 1 file changed, 9 deletions(-) (limited to 'net') diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 50deb01..3d3a094 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -321,15 +321,6 @@ static int ieee80211_open(struct net_device *dev) ieee80211_recalc_ps(local, -1); - /* - * ieee80211_sta_work is disabled while network interface - * is down. Therefore, some configuration changes may not - * yet be effective. Trigger execution of ieee80211_sta_work - * to fix this. - */ - if (sdata->vif.type == NL80211_IFTYPE_STATION) - ieee80211_queue_work(&local->hw, &sdata->u.mgd.work); - netif_tx_start_all_queues(dev); return 0; -- cgit v1.1 From 08daecaead42b85b69b33d7d8429a93dfbf75b58 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Sun, 30 May 2010 14:53:43 +0200 Subject: mac80211: drop control frames after processing After ieee80211_rx_h_ctrl() processing we only want to process management (including action) frames, so there's no point in letting control frames continue. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/rx.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'net') diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 0ade267..0b9898a 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -1844,7 +1844,12 @@ ieee80211_rx_h_ctrl(struct ieee80211_rx_data *rx, struct sk_buff_head *frames) return RX_QUEUED; } - return RX_CONTINUE; + /* + * After this point, we only want management frames, + * so we can drop all remaining control frames to + * cooked monitor interfaces. + */ + return RX_DROP_MONITOR; } static void ieee80211_process_sa_query_req(struct ieee80211_sub_if_data *sdata, -- cgit v1.1 From fcea60070fe8fa48df579f155ec7bc20a868f7dc Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 31 May 2010 11:40:23 +0200 Subject: mac80211: move plink state For some odd reason, the plink_state enum is declared in the middle between aggregation related structures. Move it down to make the file easier to read. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/sta_info.h | 44 ++++++++++++++++++++++---------------------- 1 file changed, 22 insertions(+), 22 deletions(-) (limited to 'net') diff --git a/net/mac80211/sta_info.h b/net/mac80211/sta_info.h index df9d455..813da34 100644 --- a/net/mac80211/sta_info.h +++ b/net/mac80211/sta_info.h @@ -120,6 +120,28 @@ struct tid_ampdu_rx { }; /** + * struct sta_ampdu_mlme - STA aggregation information. + * + * @tid_active_rx: TID's state in Rx session state machine. + * @tid_rx: aggregation info for Rx per TID + * @tid_state_tx: TID's state in Tx session state machine. + * @tid_tx: aggregation info for Tx per TID + * @addba_req_num: number of times addBA request has been sent. + * @dialog_token_allocator: dialog token enumerator for each new session; + */ +struct sta_ampdu_mlme { + /* rx */ + bool tid_active_rx[STA_TID_NUM]; + struct tid_ampdu_rx *tid_rx[STA_TID_NUM]; + /* tx */ + u8 tid_state_tx[STA_TID_NUM]; + struct tid_ampdu_tx *tid_tx[STA_TID_NUM]; + u8 addba_req_num[STA_TID_NUM]; + u8 dialog_token_allocator; +}; + + +/** * enum plink_state - state of a mesh peer link finite state machine * * @PLINK_LISTEN: initial state, considered the implicit state of non existant @@ -143,28 +165,6 @@ enum plink_state { }; /** - * struct sta_ampdu_mlme - STA aggregation information. - * - * @tid_active_rx: TID's state in Rx session state machine. - * @tid_rx: aggregation info for Rx per TID - * @tid_state_tx: TID's state in Tx session state machine. - * @tid_tx: aggregation info for Tx per TID - * @addba_req_num: number of times addBA request has been sent. - * @dialog_token_allocator: dialog token enumerator for each new session; - */ -struct sta_ampdu_mlme { - /* rx */ - bool tid_active_rx[STA_TID_NUM]; - struct tid_ampdu_rx *tid_rx[STA_TID_NUM]; - /* tx */ - u8 tid_state_tx[STA_TID_NUM]; - struct tid_ampdu_tx *tid_tx[STA_TID_NUM]; - u8 addba_req_num[STA_TID_NUM]; - u8 dialog_token_allocator; -}; - - -/** * struct sta_info - STA information * * This structure collects information about a station that -- cgit v1.1 From ad0e2b5a00dbec303e4682b403bb6703d11dcdb2 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 1 Jun 2010 10:19:19 +0200 Subject: mac80211: simplify key locking Since I recently made station management able to sleep, I can now rework key management as well; since it will no longer need a spinlock and can also use a mutex instead, a bunch of code to allow drivers' set_key to sleep while key management is protected by a spinlock can now be removed. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/cfg.c | 18 ++- net/mac80211/ieee80211_i.h | 4 +- net/mac80211/iface.c | 5 +- net/mac80211/key.c | 288 +++++++++++---------------------------------- net/mac80211/key.h | 22 +--- net/mac80211/main.c | 2 +- net/mac80211/sta_info.c | 8 -- 7 files changed, 80 insertions(+), 267 deletions(-) (limited to 'net') diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index f8c49c5..952845e 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -120,6 +120,9 @@ static int ieee80211_add_key(struct wiphy *wiphy, struct net_device *dev, struct ieee80211_key *key; int err; + if (!netif_running(dev)) + return -ENETDOWN; + sdata = IEEE80211_DEV_TO_SUB_IF(dev); switch (params->cipher) { @@ -145,7 +148,7 @@ static int ieee80211_add_key(struct wiphy *wiphy, struct net_device *dev, if (!key) return -ENOMEM; - rcu_read_lock(); + mutex_lock(&sdata->local->sta_mtx); if (mac_addr) { sta = sta_info_get_bss(sdata, mac_addr); @@ -160,7 +163,7 @@ static int ieee80211_add_key(struct wiphy *wiphy, struct net_device *dev, err = 0; out_unlock: - rcu_read_unlock(); + mutex_unlock(&sdata->local->sta_mtx); return err; } @@ -174,7 +177,7 @@ static int ieee80211_del_key(struct wiphy *wiphy, struct net_device *dev, sdata = IEEE80211_DEV_TO_SUB_IF(dev); - rcu_read_lock(); + mutex_lock(&sdata->local->sta_mtx); if (mac_addr) { ret = -ENOENT; @@ -202,7 +205,7 @@ static int ieee80211_del_key(struct wiphy *wiphy, struct net_device *dev, ret = 0; out_unlock: - rcu_read_unlock(); + mutex_unlock(&sdata->local->sta_mtx); return ret; } @@ -305,15 +308,10 @@ static int ieee80211_config_default_key(struct wiphy *wiphy, struct net_device *dev, u8 key_idx) { - struct ieee80211_sub_if_data *sdata; - - rcu_read_lock(); + struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); - sdata = IEEE80211_DEV_TO_SUB_IF(dev); ieee80211_set_default_key(sdata, key_idx); - rcu_read_unlock(); - return 0; } diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 47d6753..4d3883e 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -746,10 +746,10 @@ struct ieee80211_local { struct mutex iflist_mtx; /* - * Key lock, protects sdata's key_list and sta_info's + * Key mutex, protects sdata's key_list and sta_info's * key pointers (write access, they're RCU.) */ - spinlock_t key_lock; + struct mutex key_mtx; /* Scanning and BSS list */ diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 3d3a094..1afa9ec 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -268,7 +268,6 @@ static int ieee80211_open(struct net_device *dev) changed |= ieee80211_reset_erp_info(sdata); ieee80211_bss_info_change_notify(sdata, changed); - ieee80211_enable_keys(sdata); if (sdata->vif.type == NL80211_IFTYPE_STATION) netif_carrier_off(dev); @@ -522,8 +521,8 @@ static int ieee80211_stop(struct net_device *dev) BSS_CHANGED_BEACON_ENABLED); } - /* disable all keys for as long as this netdev is down */ - ieee80211_disable_keys(sdata); + /* free all remaining keys, there shouldn't be any */ + ieee80211_free_keys(sdata); drv_remove_interface(local, &sdata->vif); } diff --git a/net/mac80211/key.c b/net/mac80211/key.c index e8f6e3b..d0d9001 100644 --- a/net/mac80211/key.c +++ b/net/mac80211/key.c @@ -36,80 +36,20 @@ * There is currently no way of knowing this except by looking into * debugfs. * - * All key operations are protected internally so you can call them at - * any time. + * All key operations are protected internally. * * Within mac80211, key references are, just as STA structure references, * protected by RCU. Note, however, that some things are unprotected, * namely the key->sta dereferences within the hardware acceleration - * functions. This means that sta_info_destroy() must flush the key todo - * list. - * - * All the direct key list manipulation functions must not sleep because - * they can operate on STA info structs that are protected by RCU. + * functions. This means that sta_info_destroy() must remove the key + * which waits for an RCU grace period. */ static const u8 bcast_addr[ETH_ALEN] = { 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }; -/* key mutex: used to synchronise todo runners */ -static DEFINE_MUTEX(key_mutex); -static DEFINE_SPINLOCK(todo_lock); -static LIST_HEAD(todo_list); - -static void key_todo(struct work_struct *work) +static void assert_key_lock(struct ieee80211_local *local) { - ieee80211_key_todo(); -} - -static DECLARE_WORK(todo_work, key_todo); - -/** - * add_todo - add todo item for a key - * - * @key: key to add to do item for - * @flag: todo flag(s) - * - * Must be called with IRQs or softirqs disabled. - */ -static void add_todo(struct ieee80211_key *key, u32 flag) -{ - if (!key) - return; - - spin_lock(&todo_lock); - key->flags |= flag; - /* - * Remove again if already on the list so that we move it to the end. - */ - if (!list_empty(&key->todo)) - list_del(&key->todo); - list_add_tail(&key->todo, &todo_list); - schedule_work(&todo_work); - spin_unlock(&todo_lock); -} - -/** - * ieee80211_key_lock - lock the mac80211 key operation lock - * - * This locks the (global) mac80211 key operation lock, all - * key operations must be done under this lock. - */ -static void ieee80211_key_lock(void) -{ - mutex_lock(&key_mutex); -} - -/** - * ieee80211_key_unlock - unlock the mac80211 key operation lock - */ -static void ieee80211_key_unlock(void) -{ - mutex_unlock(&key_mutex); -} - -static void assert_key_lock(void) -{ - WARN_ON(!mutex_is_locked(&key_mutex)); + WARN_ON(!mutex_is_locked(&local->key_mtx)); } static struct ieee80211_sta *get_sta_for_key(struct ieee80211_key *key) @@ -126,12 +66,13 @@ static void ieee80211_key_enable_hw_accel(struct ieee80211_key *key) struct ieee80211_sta *sta; int ret; - assert_key_lock(); might_sleep(); if (!key->local->ops->set_key) return; + assert_key_lock(key->local); + sta = get_sta_for_key(key); sdata = key->sdata; @@ -142,11 +83,8 @@ static void ieee80211_key_enable_hw_accel(struct ieee80211_key *key) ret = drv_set_key(key->local, SET_KEY, sdata, sta, &key->conf); - if (!ret) { - spin_lock_bh(&todo_lock); + if (!ret) key->flags |= KEY_FLAG_UPLOADED_TO_HARDWARE; - spin_unlock_bh(&todo_lock); - } if (ret && ret != -ENOSPC && ret != -EOPNOTSUPP) printk(KERN_ERR "mac80211-%s: failed to set key " @@ -161,18 +99,15 @@ static void ieee80211_key_disable_hw_accel(struct ieee80211_key *key) struct ieee80211_sta *sta; int ret; - assert_key_lock(); might_sleep(); if (!key || !key->local->ops->set_key) return; - spin_lock_bh(&todo_lock); - if (!(key->flags & KEY_FLAG_UPLOADED_TO_HARDWARE)) { - spin_unlock_bh(&todo_lock); + assert_key_lock(key->local); + + if (!(key->flags & KEY_FLAG_UPLOADED_TO_HARDWARE)) return; - } - spin_unlock_bh(&todo_lock); sta = get_sta_for_key(key); sdata = key->sdata; @@ -191,9 +126,7 @@ static void ieee80211_key_disable_hw_accel(struct ieee80211_key *key) wiphy_name(key->local->hw.wiphy), key->conf.keyidx, sta ? sta->addr : bcast_addr, ret); - spin_lock_bh(&todo_lock); key->flags &= ~KEY_FLAG_UPLOADED_TO_HARDWARE; - spin_unlock_bh(&todo_lock); } static void __ieee80211_set_default_key(struct ieee80211_sub_if_data *sdata, @@ -201,22 +134,24 @@ static void __ieee80211_set_default_key(struct ieee80211_sub_if_data *sdata, { struct ieee80211_key *key = NULL; + assert_key_lock(sdata->local); + if (idx >= 0 && idx < NUM_DEFAULT_KEYS) key = sdata->keys[idx]; rcu_assign_pointer(sdata->default_key, key); - if (key) - add_todo(key, KEY_FLAG_TODO_DEFKEY); + if (key) { + ieee80211_debugfs_key_remove_default(key->sdata); + ieee80211_debugfs_key_add_default(key->sdata); + } } void ieee80211_set_default_key(struct ieee80211_sub_if_data *sdata, int idx) { - unsigned long flags; - - spin_lock_irqsave(&sdata->local->key_lock, flags); + mutex_lock(&sdata->local->key_mtx); __ieee80211_set_default_key(sdata, idx); - spin_unlock_irqrestore(&sdata->local->key_lock, flags); + mutex_unlock(&sdata->local->key_mtx); } static void @@ -224,24 +159,26 @@ __ieee80211_set_default_mgmt_key(struct ieee80211_sub_if_data *sdata, int idx) { struct ieee80211_key *key = NULL; + assert_key_lock(sdata->local); + if (idx >= NUM_DEFAULT_KEYS && idx < NUM_DEFAULT_KEYS + NUM_DEFAULT_MGMT_KEYS) key = sdata->keys[idx]; rcu_assign_pointer(sdata->default_mgmt_key, key); - if (key) - add_todo(key, KEY_FLAG_TODO_DEFMGMTKEY); + if (key) { + ieee80211_debugfs_key_remove_mgmt_default(key->sdata); + ieee80211_debugfs_key_add_mgmt_default(key->sdata); + } } void ieee80211_set_default_mgmt_key(struct ieee80211_sub_if_data *sdata, int idx) { - unsigned long flags; - - spin_lock_irqsave(&sdata->local->key_lock, flags); + mutex_lock(&sdata->local->key_mtx); __ieee80211_set_default_mgmt_key(sdata, idx); - spin_unlock_irqrestore(&sdata->local->key_lock, flags); + mutex_unlock(&sdata->local->key_mtx); } @@ -352,7 +289,6 @@ struct ieee80211_key *ieee80211_key_alloc(enum ieee80211_key_alg alg, } memcpy(key->conf.key, key_data, key_len); INIT_LIST_HEAD(&key->list); - INIT_LIST_HEAD(&key->todo); if (alg == ALG_CCMP) { /* @@ -382,12 +318,27 @@ struct ieee80211_key *ieee80211_key_alloc(enum ieee80211_key_alg alg, return key; } +static void __ieee80211_key_destroy(struct ieee80211_key *key) +{ + if (!key) + return; + + ieee80211_key_disable_hw_accel(key); + + if (key->conf.alg == ALG_CCMP) + ieee80211_aes_key_free(key->u.ccmp.tfm); + if (key->conf.alg == ALG_AES_CMAC) + ieee80211_aes_cmac_key_free(key->u.aes_cmac.tfm); + ieee80211_debugfs_key_remove(key); + + kfree(key); +} + void ieee80211_key_link(struct ieee80211_key *key, struct ieee80211_sub_if_data *sdata, struct sta_info *sta) { struct ieee80211_key *old_key; - unsigned long flags; int idx; BUG_ON(!sdata); @@ -431,7 +382,7 @@ void ieee80211_key_link(struct ieee80211_key *key, } } - spin_lock_irqsave(&sdata->local->key_lock, flags); + mutex_lock(&sdata->local->key_mtx); if (sta) old_key = sta->key; @@ -439,15 +390,13 @@ void ieee80211_key_link(struct ieee80211_key *key, old_key = sdata->keys[idx]; __ieee80211_key_replace(sdata, sta, old_key, key); + __ieee80211_key_destroy(old_key); - /* free old key later */ - add_todo(old_key, KEY_FLAG_TODO_DELETE); + ieee80211_debugfs_key_add(key); - add_todo(key, KEY_FLAG_TODO_ADD_DEBUGFS); - if (ieee80211_sdata_running(sdata)) - add_todo(key, KEY_FLAG_TODO_HWACCEL_ADD); + ieee80211_key_enable_hw_accel(key); - spin_unlock_irqrestore(&sdata->local->key_lock, flags); + mutex_unlock(&sdata->local->key_mtx); } static void __ieee80211_key_free(struct ieee80211_key *key) @@ -458,170 +407,65 @@ static void __ieee80211_key_free(struct ieee80211_key *key) if (key->sdata) __ieee80211_key_replace(key->sdata, key->sta, key, NULL); - - add_todo(key, KEY_FLAG_TODO_DELETE); + __ieee80211_key_destroy(key); } void ieee80211_key_free(struct ieee80211_key *key) { - unsigned long flags; + struct ieee80211_local *local; if (!key) return; - if (!key->sdata) { - /* The key has not been linked yet, simply free it - * and don't Oops */ - if (key->conf.alg == ALG_CCMP) - ieee80211_aes_key_free(key->u.ccmp.tfm); - kfree(key); - return; - } + local = key->sdata->local; - spin_lock_irqsave(&key->sdata->local->key_lock, flags); + mutex_lock(&local->key_mtx); __ieee80211_key_free(key); - spin_unlock_irqrestore(&key->sdata->local->key_lock, flags); + mutex_unlock(&local->key_mtx); } -/* - * To be safe against concurrent manipulations of the list (which shouldn't - * actually happen) we need to hold the spinlock. But under the spinlock we - * can't actually do much, so we defer processing to the todo list. Then run - * the todo list to be sure the operation and possibly previously pending - * operations are completed. - */ -static void ieee80211_todo_for_each_key(struct ieee80211_sub_if_data *sdata, - u32 todo_flags) +void ieee80211_enable_keys(struct ieee80211_sub_if_data *sdata) { struct ieee80211_key *key; - unsigned long flags; - - might_sleep(); - - spin_lock_irqsave(&sdata->local->key_lock, flags); - list_for_each_entry(key, &sdata->key_list, list) - add_todo(key, todo_flags); - spin_unlock_irqrestore(&sdata->local->key_lock, flags); - - ieee80211_key_todo(); -} -void ieee80211_enable_keys(struct ieee80211_sub_if_data *sdata) -{ ASSERT_RTNL(); if (WARN_ON(!ieee80211_sdata_running(sdata))) return; - ieee80211_todo_for_each_key(sdata, KEY_FLAG_TODO_HWACCEL_ADD); -} - -void ieee80211_disable_keys(struct ieee80211_sub_if_data *sdata) -{ - ASSERT_RTNL(); - - ieee80211_todo_for_each_key(sdata, KEY_FLAG_TODO_HWACCEL_REMOVE); -} - -static void __ieee80211_key_destroy(struct ieee80211_key *key) -{ - if (!key) - return; - - ieee80211_key_disable_hw_accel(key); + mutex_lock(&sdata->local->key_mtx); - if (key->conf.alg == ALG_CCMP) - ieee80211_aes_key_free(key->u.ccmp.tfm); - if (key->conf.alg == ALG_AES_CMAC) - ieee80211_aes_cmac_key_free(key->u.aes_cmac.tfm); - ieee80211_debugfs_key_remove(key); + list_for_each_entry(key, &sdata->key_list, list) + ieee80211_key_enable_hw_accel(key); - kfree(key); + mutex_unlock(&sdata->local->key_mtx); } -static void __ieee80211_key_todo(void) +void ieee80211_disable_keys(struct ieee80211_sub_if_data *sdata) { struct ieee80211_key *key; - bool work_done; - u32 todoflags; - /* - * NB: sta_info_destroy relies on this! - */ - synchronize_rcu(); - - spin_lock_bh(&todo_lock); - while (!list_empty(&todo_list)) { - key = list_first_entry(&todo_list, struct ieee80211_key, todo); - list_del_init(&key->todo); - todoflags = key->flags & (KEY_FLAG_TODO_ADD_DEBUGFS | - KEY_FLAG_TODO_DEFKEY | - KEY_FLAG_TODO_DEFMGMTKEY | - KEY_FLAG_TODO_HWACCEL_ADD | - KEY_FLAG_TODO_HWACCEL_REMOVE | - KEY_FLAG_TODO_DELETE); - key->flags &= ~todoflags; - spin_unlock_bh(&todo_lock); - - work_done = false; - - if (todoflags & KEY_FLAG_TODO_ADD_DEBUGFS) { - ieee80211_debugfs_key_add(key); - work_done = true; - } - if (todoflags & KEY_FLAG_TODO_DEFKEY) { - ieee80211_debugfs_key_remove_default(key->sdata); - ieee80211_debugfs_key_add_default(key->sdata); - work_done = true; - } - if (todoflags & KEY_FLAG_TODO_DEFMGMTKEY) { - ieee80211_debugfs_key_remove_mgmt_default(key->sdata); - ieee80211_debugfs_key_add_mgmt_default(key->sdata); - work_done = true; - } - if (todoflags & KEY_FLAG_TODO_HWACCEL_ADD) { - ieee80211_key_enable_hw_accel(key); - work_done = true; - } - if (todoflags & KEY_FLAG_TODO_HWACCEL_REMOVE) { - ieee80211_key_disable_hw_accel(key); - work_done = true; - } - if (todoflags & KEY_FLAG_TODO_DELETE) { - __ieee80211_key_destroy(key); - work_done = true; - } + ASSERT_RTNL(); - WARN_ON(!work_done); + mutex_lock(&sdata->local->key_mtx); - spin_lock_bh(&todo_lock); - } - spin_unlock_bh(&todo_lock); -} + list_for_each_entry(key, &sdata->key_list, list) + ieee80211_key_disable_hw_accel(key); -void ieee80211_key_todo(void) -{ - ieee80211_key_lock(); - __ieee80211_key_todo(); - ieee80211_key_unlock(); + mutex_unlock(&sdata->local->key_mtx); } void ieee80211_free_keys(struct ieee80211_sub_if_data *sdata) { struct ieee80211_key *key, *tmp; - unsigned long flags; - ieee80211_key_lock(); + mutex_lock(&sdata->local->key_mtx); ieee80211_debugfs_key_remove_default(sdata); ieee80211_debugfs_key_remove_mgmt_default(sdata); - spin_lock_irqsave(&sdata->local->key_lock, flags); list_for_each_entry_safe(key, tmp, &sdata->key_list, list) __ieee80211_key_free(key); - spin_unlock_irqrestore(&sdata->local->key_lock, flags); - - __ieee80211_key_todo(); - ieee80211_key_unlock(); + mutex_unlock(&sdata->local->key_mtx); } diff --git a/net/mac80211/key.h b/net/mac80211/key.h index bdc2968..9996e3b 100644 --- a/net/mac80211/key.h +++ b/net/mac80211/key.h @@ -38,25 +38,9 @@ struct sta_info; * * @KEY_FLAG_UPLOADED_TO_HARDWARE: Indicates that this key is present * in the hardware for TX crypto hardware acceleration. - * @KEY_FLAG_TODO_DELETE: Key is marked for deletion and will, after an - * RCU grace period, no longer be reachable other than from the - * todo list. - * @KEY_FLAG_TODO_HWACCEL_ADD: Key needs to be added to hardware acceleration. - * @KEY_FLAG_TODO_HWACCEL_REMOVE: Key needs to be removed from hardware - * acceleration. - * @KEY_FLAG_TODO_DEFKEY: Key is default key and debugfs needs to be updated. - * @KEY_FLAG_TODO_ADD_DEBUGFS: Key needs to be added to debugfs. - * @KEY_FLAG_TODO_DEFMGMTKEY: Key is default management key and debugfs needs - * to be updated. */ enum ieee80211_internal_key_flags { KEY_FLAG_UPLOADED_TO_HARDWARE = BIT(0), - KEY_FLAG_TODO_DELETE = BIT(1), - KEY_FLAG_TODO_HWACCEL_ADD = BIT(2), - KEY_FLAG_TODO_HWACCEL_REMOVE = BIT(3), - KEY_FLAG_TODO_DEFKEY = BIT(4), - KEY_FLAG_TODO_ADD_DEBUGFS = BIT(5), - KEY_FLAG_TODO_DEFMGMTKEY = BIT(6), }; enum ieee80211_internal_tkip_state { @@ -79,10 +63,8 @@ struct ieee80211_key { /* for sdata list */ struct list_head list; - /* for todo list */ - struct list_head todo; - /* protected by todo lock! */ + /* protected by key mutex */ unsigned int flags; union { @@ -155,6 +137,4 @@ void ieee80211_free_keys(struct ieee80211_sub_if_data *sdata); void ieee80211_enable_keys(struct ieee80211_sub_if_data *sdata); void ieee80211_disable_keys(struct ieee80211_sub_if_data *sdata); -void ieee80211_key_todo(void); - #endif /* IEEE80211_KEY_H */ diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 4051b23..045ead9 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -448,7 +448,7 @@ struct ieee80211_hw *ieee80211_alloc_hw(size_t priv_data_len, mutex_init(&local->iflist_mtx); mutex_init(&local->scan_mtx); - spin_lock_init(&local->key_lock); + mutex_init(&local->key_mtx); spin_lock_init(&local->filter_lock); spin_lock_init(&local->queue_stop_reason_lock); diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c index 7301975..c426c57 100644 --- a/net/mac80211/sta_info.c +++ b/net/mac80211/sta_info.c @@ -648,14 +648,6 @@ static int __must_check __sta_info_destroy(struct sta_info *sta) if (sta->key) { ieee80211_key_free(sta->key); - /* - * We have only unlinked the key, and actually destroying it - * may mean it is removed from hardware which requires that - * the key->sta pointer is still valid, so flush the key todo - * list here. - */ - ieee80211_key_todo(); - WARN_ON(sta->key); } -- cgit v1.1 From 2826bcd844e05dcbef9b9284bddb7fe88e8d314f Mon Sep 17 00:00:00 2001 From: Felix Fietkau Date: Wed, 2 Jun 2010 02:57:34 +0200 Subject: mac80211: reduce debugfs code size This patch reduces the binary size by around 25k (measured on MIPS, with CONFIG_MAC80211_DEBUG_COUNTERS enabled). Signed-off-by: Felix Fietkau Signed-off-by: John W. Linville --- net/mac80211/debugfs.c | 154 +++++++++++++++++---------------------------- net/mac80211/debugfs_sta.c | 45 ++++++------- 2 files changed, 77 insertions(+), 122 deletions(-) (limited to 'net') diff --git a/net/mac80211/debugfs.c b/net/mac80211/debugfs.c index 637929b..a694c59 100644 --- a/net/mac80211/debugfs.c +++ b/net/mac80211/debugfs.c @@ -307,9 +307,6 @@ static const struct file_operations queues_ops = { /* statistics stuff */ -#define DEBUGFS_STATS_FILE(name, buflen, fmt, value...) \ - DEBUGFS_READONLY_FILE(stats_ ##name, buflen, fmt, ##value) - static ssize_t format_devstat_counter(struct ieee80211_local *local, char __user *userbuf, size_t count, loff_t *ppos, @@ -351,75 +348,16 @@ static const struct file_operations stats_ ##name## _ops = { \ .open = mac80211_open_file_generic, \ }; -#define DEBUGFS_STATS_ADD(name) \ +#define DEBUGFS_STATS_ADD(name, field) \ + debugfs_create_u32(#name, 0400, statsd, (u32 *) &field); +#define DEBUGFS_DEVSTATS_ADD(name) \ debugfs_create_file(#name, 0400, statsd, local, &stats_ ##name## _ops); -DEBUGFS_STATS_FILE(transmitted_fragment_count, 20, "%u", - local->dot11TransmittedFragmentCount); -DEBUGFS_STATS_FILE(multicast_transmitted_frame_count, 20, "%u", - local->dot11MulticastTransmittedFrameCount); -DEBUGFS_STATS_FILE(failed_count, 20, "%u", - local->dot11FailedCount); -DEBUGFS_STATS_FILE(retry_count, 20, "%u", - local->dot11RetryCount); -DEBUGFS_STATS_FILE(multiple_retry_count, 20, "%u", - local->dot11MultipleRetryCount); -DEBUGFS_STATS_FILE(frame_duplicate_count, 20, "%u", - local->dot11FrameDuplicateCount); -DEBUGFS_STATS_FILE(received_fragment_count, 20, "%u", - local->dot11ReceivedFragmentCount); -DEBUGFS_STATS_FILE(multicast_received_frame_count, 20, "%u", - local->dot11MulticastReceivedFrameCount); -DEBUGFS_STATS_FILE(transmitted_frame_count, 20, "%u", - local->dot11TransmittedFrameCount); -#ifdef CONFIG_MAC80211_DEBUG_COUNTERS -DEBUGFS_STATS_FILE(tx_handlers_drop, 20, "%u", - local->tx_handlers_drop); -DEBUGFS_STATS_FILE(tx_handlers_queued, 20, "%u", - local->tx_handlers_queued); -DEBUGFS_STATS_FILE(tx_handlers_drop_unencrypted, 20, "%u", - local->tx_handlers_drop_unencrypted); -DEBUGFS_STATS_FILE(tx_handlers_drop_fragment, 20, "%u", - local->tx_handlers_drop_fragment); -DEBUGFS_STATS_FILE(tx_handlers_drop_wep, 20, "%u", - local->tx_handlers_drop_wep); -DEBUGFS_STATS_FILE(tx_handlers_drop_not_assoc, 20, "%u", - local->tx_handlers_drop_not_assoc); -DEBUGFS_STATS_FILE(tx_handlers_drop_unauth_port, 20, "%u", - local->tx_handlers_drop_unauth_port); -DEBUGFS_STATS_FILE(rx_handlers_drop, 20, "%u", - local->rx_handlers_drop); -DEBUGFS_STATS_FILE(rx_handlers_queued, 20, "%u", - local->rx_handlers_queued); -DEBUGFS_STATS_FILE(rx_handlers_drop_nullfunc, 20, "%u", - local->rx_handlers_drop_nullfunc); -DEBUGFS_STATS_FILE(rx_handlers_drop_defrag, 20, "%u", - local->rx_handlers_drop_defrag); -DEBUGFS_STATS_FILE(rx_handlers_drop_short, 20, "%u", - local->rx_handlers_drop_short); -DEBUGFS_STATS_FILE(rx_handlers_drop_passive_scan, 20, "%u", - local->rx_handlers_drop_passive_scan); -DEBUGFS_STATS_FILE(tx_expand_skb_head, 20, "%u", - local->tx_expand_skb_head); -DEBUGFS_STATS_FILE(tx_expand_skb_head_cloned, 20, "%u", - local->tx_expand_skb_head_cloned); -DEBUGFS_STATS_FILE(rx_expand_skb_head, 20, "%u", - local->rx_expand_skb_head); -DEBUGFS_STATS_FILE(rx_expand_skb_head2, 20, "%u", - local->rx_expand_skb_head2); -DEBUGFS_STATS_FILE(rx_handlers_fragments, 20, "%u", - local->rx_handlers_fragments); -DEBUGFS_STATS_FILE(tx_status_drop, 20, "%u", - local->tx_status_drop); - -#endif - DEBUGFS_DEVSTATS_FILE(dot11ACKFailureCount); DEBUGFS_DEVSTATS_FILE(dot11RTSFailureCount); DEBUGFS_DEVSTATS_FILE(dot11FCSErrorCount); DEBUGFS_DEVSTATS_FILE(dot11RTSSuccessCount); - void debugfs_hw_add(struct ieee80211_local *local) { struct dentry *phyd = local->hw.wiphy->debugfsdir; @@ -448,38 +386,60 @@ void debugfs_hw_add(struct ieee80211_local *local) if (!statsd) return; - DEBUGFS_STATS_ADD(transmitted_fragment_count); - DEBUGFS_STATS_ADD(multicast_transmitted_frame_count); - DEBUGFS_STATS_ADD(failed_count); - DEBUGFS_STATS_ADD(retry_count); - DEBUGFS_STATS_ADD(multiple_retry_count); - DEBUGFS_STATS_ADD(frame_duplicate_count); - DEBUGFS_STATS_ADD(received_fragment_count); - DEBUGFS_STATS_ADD(multicast_received_frame_count); - DEBUGFS_STATS_ADD(transmitted_frame_count); + DEBUGFS_STATS_ADD(transmitted_fragment_count, + local->dot11TransmittedFragmentCount); + DEBUGFS_STATS_ADD(multicast_transmitted_frame_count, + local->dot11MulticastTransmittedFrameCount); + DEBUGFS_STATS_ADD(failed_count, local->dot11FailedCount); + DEBUGFS_STATS_ADD(retry_count, local->dot11RetryCount); + DEBUGFS_STATS_ADD(multiple_retry_count, + local->dot11MultipleRetryCount); + DEBUGFS_STATS_ADD(frame_duplicate_count, + local->dot11FrameDuplicateCount); + DEBUGFS_STATS_ADD(received_fragment_count, + local->dot11ReceivedFragmentCount); + DEBUGFS_STATS_ADD(multicast_received_frame_count, + local->dot11MulticastReceivedFrameCount); + DEBUGFS_STATS_ADD(transmitted_frame_count, + local->dot11TransmittedFrameCount); #ifdef CONFIG_MAC80211_DEBUG_COUNTERS - DEBUGFS_STATS_ADD(tx_handlers_drop); - DEBUGFS_STATS_ADD(tx_handlers_queued); - DEBUGFS_STATS_ADD(tx_handlers_drop_unencrypted); - DEBUGFS_STATS_ADD(tx_handlers_drop_fragment); - DEBUGFS_STATS_ADD(tx_handlers_drop_wep); - DEBUGFS_STATS_ADD(tx_handlers_drop_not_assoc); - DEBUGFS_STATS_ADD(tx_handlers_drop_unauth_port); - DEBUGFS_STATS_ADD(rx_handlers_drop); - DEBUGFS_STATS_ADD(rx_handlers_queued); - DEBUGFS_STATS_ADD(rx_handlers_drop_nullfunc); - DEBUGFS_STATS_ADD(rx_handlers_drop_defrag); - DEBUGFS_STATS_ADD(rx_handlers_drop_short); - DEBUGFS_STATS_ADD(rx_handlers_drop_passive_scan); - DEBUGFS_STATS_ADD(tx_expand_skb_head); - DEBUGFS_STATS_ADD(tx_expand_skb_head_cloned); - DEBUGFS_STATS_ADD(rx_expand_skb_head); - DEBUGFS_STATS_ADD(rx_expand_skb_head2); - DEBUGFS_STATS_ADD(rx_handlers_fragments); - DEBUGFS_STATS_ADD(tx_status_drop); + DEBUGFS_STATS_ADD(tx_handlers_drop, local->tx_handlers_drop); + DEBUGFS_STATS_ADD(tx_handlers_queued, local->tx_handlers_queued); + DEBUGFS_STATS_ADD(tx_handlers_drop_unencrypted, + local->tx_handlers_drop_unencrypted); + DEBUGFS_STATS_ADD(tx_handlers_drop_fragment, + local->tx_handlers_drop_fragment); + DEBUGFS_STATS_ADD(tx_handlers_drop_wep, + local->tx_handlers_drop_wep); + DEBUGFS_STATS_ADD(tx_handlers_drop_not_assoc, + local->tx_handlers_drop_not_assoc); + DEBUGFS_STATS_ADD(tx_handlers_drop_unauth_port, + local->tx_handlers_drop_unauth_port); + DEBUGFS_STATS_ADD(rx_handlers_drop, local->rx_handlers_drop); + DEBUGFS_STATS_ADD(rx_handlers_queued, local->rx_handlers_queued); + DEBUGFS_STATS_ADD(rx_handlers_drop_nullfunc, + local->rx_handlers_drop_nullfunc); + DEBUGFS_STATS_ADD(rx_handlers_drop_defrag, + local->rx_handlers_drop_defrag); + DEBUGFS_STATS_ADD(rx_handlers_drop_short, + local->rx_handlers_drop_short); + DEBUGFS_STATS_ADD(rx_handlers_drop_passive_scan, + local->rx_handlers_drop_passive_scan); + DEBUGFS_STATS_ADD(tx_expand_skb_head, + local->tx_expand_skb_head); + DEBUGFS_STATS_ADD(tx_expand_skb_head_cloned, + local->tx_expand_skb_head_cloned); + DEBUGFS_STATS_ADD(rx_expand_skb_head, + local->rx_expand_skb_head); + DEBUGFS_STATS_ADD(rx_expand_skb_head2, + local->rx_expand_skb_head2); + DEBUGFS_STATS_ADD(rx_handlers_fragments, + local->rx_handlers_fragments); + DEBUGFS_STATS_ADD(tx_status_drop, + local->tx_status_drop); #endif - DEBUGFS_STATS_ADD(dot11ACKFailureCount); - DEBUGFS_STATS_ADD(dot11RTSFailureCount); - DEBUGFS_STATS_ADD(dot11FCSErrorCount); - DEBUGFS_STATS_ADD(dot11RTSSuccessCount); + DEBUGFS_DEVSTATS_ADD(dot11ACKFailureCount); + DEBUGFS_DEVSTATS_ADD(dot11RTSFailureCount); + DEBUGFS_DEVSTATS_ADD(dot11FCSErrorCount); + DEBUGFS_DEVSTATS_ADD(dot11RTSSuccessCount); } diff --git a/net/mac80211/debugfs_sta.c b/net/mac80211/debugfs_sta.c index 9f14061..576e024 100644 --- a/net/mac80211/debugfs_sta.c +++ b/net/mac80211/debugfs_sta.c @@ -30,7 +30,6 @@ static ssize_t sta_ ##name## _read(struct file *file, \ } #define STA_READ_D(name, field) STA_READ(name, 20, field, "%d\n") #define STA_READ_U(name, field) STA_READ(name, 20, field, "%u\n") -#define STA_READ_LU(name, field) STA_READ(name, 20, field, "%lu\n") #define STA_READ_S(name, field) STA_READ(name, 20, field, "%s\n") #define STA_OPS(name) \ @@ -52,19 +51,7 @@ static const struct file_operations sta_ ##name## _ops = { \ STA_FILE(aid, sta.aid, D); STA_FILE(dev, sdata->name, S); -STA_FILE(rx_packets, rx_packets, LU); -STA_FILE(tx_packets, tx_packets, LU); -STA_FILE(rx_bytes, rx_bytes, LU); -STA_FILE(tx_bytes, tx_bytes, LU); -STA_FILE(rx_duplicates, num_duplicates, LU); -STA_FILE(rx_fragments, rx_fragments, LU); -STA_FILE(rx_dropped, rx_dropped, LU); -STA_FILE(tx_fragments, tx_fragments, LU); -STA_FILE(tx_filtered, tx_filtered_count, LU); -STA_FILE(tx_retry_failed, tx_retry_failed, LU); -STA_FILE(tx_retry_count, tx_retry_count, LU); STA_FILE(last_signal, last_signal, D); -STA_FILE(wep_weak_iv_count, wep_weak_iv_count, LU); static ssize_t sta_flags_read(struct file *file, char __user *userbuf, size_t count, loff_t *ppos) @@ -306,6 +293,13 @@ STA_OPS(ht_capa); debugfs_create_file(#name, 0400, \ sta->debugfs.dir, sta, &sta_ ##name## _ops); +#define DEBUGFS_ADD_COUNTER(name, field) \ + if (sizeof(sta->field) == sizeof(u32)) \ + debugfs_create_u32(#name, 0400, sta->debugfs.dir, \ + (u32 *) &sta->field); \ + else \ + debugfs_create_u64(#name, 0400, sta->debugfs.dir, \ + (u64 *) &sta->field); void ieee80211_sta_debugfs_add(struct sta_info *sta) { @@ -338,20 +332,21 @@ void ieee80211_sta_debugfs_add(struct sta_info *sta) DEBUGFS_ADD(last_seq_ctrl); DEBUGFS_ADD(agg_status); DEBUGFS_ADD(dev); - DEBUGFS_ADD(rx_packets); - DEBUGFS_ADD(tx_packets); - DEBUGFS_ADD(rx_bytes); - DEBUGFS_ADD(tx_bytes); - DEBUGFS_ADD(rx_duplicates); - DEBUGFS_ADD(rx_fragments); - DEBUGFS_ADD(rx_dropped); - DEBUGFS_ADD(tx_fragments); - DEBUGFS_ADD(tx_filtered); - DEBUGFS_ADD(tx_retry_failed); - DEBUGFS_ADD(tx_retry_count); DEBUGFS_ADD(last_signal); - DEBUGFS_ADD(wep_weak_iv_count); DEBUGFS_ADD(ht_capa); + + DEBUGFS_ADD_COUNTER(rx_packets, rx_packets); + DEBUGFS_ADD_COUNTER(tx_packets, tx_packets); + DEBUGFS_ADD_COUNTER(rx_bytes, rx_bytes); + DEBUGFS_ADD_COUNTER(tx_bytes, tx_bytes); + DEBUGFS_ADD_COUNTER(rx_duplicates, num_duplicates); + DEBUGFS_ADD_COUNTER(rx_fragments, rx_fragments); + DEBUGFS_ADD_COUNTER(rx_dropped, rx_dropped); + DEBUGFS_ADD_COUNTER(tx_fragments, tx_fragments); + DEBUGFS_ADD_COUNTER(tx_filtered, tx_filtered_count); + DEBUGFS_ADD_COUNTER(tx_retry_failed, tx_retry_failed); + DEBUGFS_ADD_COUNTER(tx_retry_count, tx_retry_count); + DEBUGFS_ADD_COUNTER(wep_weak_iv_count, wep_weak_iv_count); } void ieee80211_sta_debugfs_remove(struct sta_info *sta) -- cgit v1.1 From 4736022844fe694c4ee971fa2b6c1cb38dadbc78 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 3 Jun 2010 04:13:21 +0000 Subject: ipv4: RCU changes in __mkroute_input() Avoid two atomic ops on output device refcount Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/route.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) (limited to 'net') diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 1cfe0d1..7b8eacd 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -1957,22 +1957,22 @@ static void ip_handle_martian_source(struct net_device *dev, #endif } +/* called in rcu_read_lock() section */ static int __mkroute_input(struct sk_buff *skb, struct fib_result *res, struct in_device *in_dev, __be32 daddr, __be32 saddr, u32 tos, struct rtable **result) { - struct rtable *rth; int err; struct in_device *out_dev; - unsigned flags = 0; + unsigned int flags = 0; __be32 spec_dst; u32 itag; /* get a working reference to the output device */ - out_dev = in_dev_get(FIB_RES_DEV(*res)); + out_dev = __in_dev_get_rcu(FIB_RES_DEV(*res)); if (out_dev == NULL) { if (net_ratelimit()) printk(KERN_CRIT "Bug in ip_route_input" \ @@ -2053,8 +2053,6 @@ static int __mkroute_input(struct sk_buff *skb, *result = rth; err = 0; cleanup: - /* release the working reference to the output device */ - in_dev_put(out_dev); return err; } -- cgit v1.1 From faa9dcf793beba05f7178b63a59eaa3ca5175b6a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 3 Jun 2010 04:09:10 +0000 Subject: arp: RCU changes Avoid two atomic ops in arp_fwd_proxy() Avoid two atomic ops in arp_process() Valid optims since arp_rcv() is run under rcu_read_lock() Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/arp.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) (limited to 'net') diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index f094b75..917d2d6 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -545,10 +545,10 @@ static inline int arp_fwd_proxy(struct in_device *in_dev, /* place to check for proxy_arp for routes */ - if ((out_dev = in_dev_get(rt->u.dst.dev)) != NULL) { + out_dev = __in_dev_get_rcu(rt->u.dst.dev); + if (out_dev) omi = IN_DEV_MEDIUM_ID(out_dev); - in_dev_put(out_dev); - } + return (omi != imi && omi != -1); } @@ -741,7 +741,7 @@ void arp_send(int type, int ptype, __be32 dest_ip, static int arp_process(struct sk_buff *skb) { struct net_device *dev = skb->dev; - struct in_device *in_dev = in_dev_get(dev); + struct in_device *in_dev = __in_dev_get_rcu(dev); struct arphdr *arp; unsigned char *arp_ptr; struct rtable *rt; @@ -890,7 +890,6 @@ static int arp_process(struct sk_buff *skb) arp_send(ARPOP_REPLY,ETH_P_ARP,sip,dev,tip,sha,dev->dev_addr,sha); } else { pneigh_enqueue(&arp_tbl, in_dev->arp_parms, skb); - in_dev_put(in_dev); return 0; } goto out; @@ -936,8 +935,6 @@ static int arp_process(struct sk_buff *skb) } out: - if (in_dev) - in_dev_put(in_dev); consume_skb(skb); return 0; } -- cgit v1.1 From c6d409cfd0fd41e7a0847875e4338ad648c9b96b Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 3 Jun 2010 20:03:40 -0700 Subject: From abbffa2aa9bd6f8df16d0d0a102af677510d8b9a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 3 Jun 2010 04:29:41 +0000 Subject: [PATCH 2/3] net: net/socket.c and net/compat.c cleanups cleanup patch, to match modern coding style. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/compat.c | 47 ++++++++--------- net/socket.c | 165 ++++++++++++++++++++++++++++------------------------------ 2 files changed, 102 insertions(+), 110 deletions(-) diff --git a/net/compat.c b/net/compat.c index 1cf7590..63d260e 100644 --- a/net/compat.c +++ b/net/compat.c @@ -81,7 +81,7 @@ int verify_compat_iovec(struct msghdr *kern_msg, struct iovec *kern_iov, int tot_len; if (kern_msg->msg_namelen) { - if (mode==VERIFY_READ) { + if (mode == VERIFY_READ) { int err = move_addr_to_kernel(kern_msg->msg_name, kern_msg->msg_namelen, kern_address); @@ -354,7 +354,7 @@ static int do_set_attach_filter(struct socket *sock, int level, int optname, static int do_set_sock_timeout(struct socket *sock, int level, int optname, char __user *optval, unsigned int optlen) { - struct compat_timeval __user *up = (struct compat_timeval __user *) optval; + struct compat_timeval __user *up = (struct compat_timeval __user *)optval; struct timeval ktime; mm_segment_t old_fs; int err; @@ -367,7 +367,7 @@ static int do_set_sock_timeout(struct socket *sock, int level, return -EFAULT; old_fs = get_fs(); set_fs(KERNEL_DS); - err = sock_setsockopt(sock, level, optname, (char *) &ktime, sizeof(ktime)); + err = sock_setsockopt(sock, level, optname, (char *)&ktime, sizeof(ktime)); set_fs(old_fs); return err; @@ -389,11 +389,10 @@ asmlinkage long compat_sys_setsockopt(int fd, int level, int optname, char __user *optval, unsigned int optlen) { int err; - struct socket *sock; + struct socket *sock = sockfd_lookup(fd, &err); - if ((sock = sockfd_lookup(fd, &err))!=NULL) - { - err = security_socket_setsockopt(sock,level,optname); + if (sock) { + err = security_socket_setsockopt(sock, level, optname); if (err) { sockfd_put(sock); return err; @@ -453,7 +452,7 @@ static int compat_sock_getsockopt(struct socket *sock, int level, int optname, int compat_sock_get_timestamp(struct sock *sk, struct timeval __user *userstamp) { struct compat_timeval __user *ctv = - (struct compat_timeval __user*) userstamp; + (struct compat_timeval __user *) userstamp; int err = -ENOENT; struct timeval tv; @@ -477,7 +476,7 @@ EXPORT_SYMBOL(compat_sock_get_timestamp); int compat_sock_get_timestampns(struct sock *sk, struct timespec __user *userstamp) { struct compat_timespec __user *ctv = - (struct compat_timespec __user*) userstamp; + (struct compat_timespec __user *) userstamp; int err = -ENOENT; struct timespec ts; @@ -502,12 +501,10 @@ asmlinkage long compat_sys_getsockopt(int fd, int level, int optname, char __user *optval, int __user *optlen) { int err; - struct socket *sock; + struct socket *sock = sockfd_lookup(fd, &err); - if ((sock = sockfd_lookup(fd, &err))!=NULL) - { - err = security_socket_getsockopt(sock, level, - optname); + if (sock) { + err = security_socket_getsockopt(sock, level, optname); if (err) { sockfd_put(sock); return err; @@ -557,7 +554,7 @@ struct compat_group_filter { int compat_mc_setsockopt(struct sock *sock, int level, int optname, char __user *optval, unsigned int optlen, - int (*setsockopt)(struct sock *,int,int,char __user *,unsigned int)) + int (*setsockopt)(struct sock *, int, int, char __user *, unsigned int)) { char __user *koptval = optval; int koptlen = optlen; @@ -640,12 +637,11 @@ int compat_mc_setsockopt(struct sock *sock, int level, int optname, } return setsockopt(sock, level, optname, koptval, koptlen); } - EXPORT_SYMBOL(compat_mc_setsockopt); int compat_mc_getsockopt(struct sock *sock, int level, int optname, char __user *optval, int __user *optlen, - int (*getsockopt)(struct sock *,int,int,char __user *,int __user *)) + int (*getsockopt)(struct sock *, int, int, char __user *, int __user *)) { struct compat_group_filter __user *gf32 = (void *)optval; struct group_filter __user *kgf; @@ -681,7 +677,7 @@ int compat_mc_getsockopt(struct sock *sock, int level, int optname, __put_user(interface, &kgf->gf_interface) || __put_user(fmode, &kgf->gf_fmode) || __put_user(numsrc, &kgf->gf_numsrc) || - copy_in_user(&kgf->gf_group,&gf32->gf_group,sizeof(kgf->gf_group))) + copy_in_user(&kgf->gf_group, &gf32->gf_group, sizeof(kgf->gf_group))) return -EFAULT; err = getsockopt(sock, level, optname, (char __user *)kgf, koptlen); @@ -714,21 +710,22 @@ int compat_mc_getsockopt(struct sock *sock, int level, int optname, copylen = numsrc * sizeof(gf32->gf_slist[0]); if (copylen > klen) copylen = klen; - if (copy_in_user(gf32->gf_slist, kgf->gf_slist, copylen)) + if (copy_in_user(gf32->gf_slist, kgf->gf_slist, copylen)) return -EFAULT; } return err; } - EXPORT_SYMBOL(compat_mc_getsockopt); /* Argument list sizes for compat_sys_socketcall */ #define AL(x) ((x) * sizeof(u32)) -static unsigned char nas[20]={AL(0),AL(3),AL(3),AL(3),AL(2),AL(3), - AL(3),AL(3),AL(4),AL(4),AL(4),AL(6), - AL(6),AL(2),AL(5),AL(5),AL(3),AL(3), - AL(4),AL(5)}; +static unsigned char nas[20] = { + AL(0), AL(3), AL(3), AL(3), AL(2), AL(3), + AL(3), AL(3), AL(4), AL(4), AL(4), AL(6), + AL(6), AL(2), AL(5), AL(5), AL(3), AL(3), + AL(4), AL(5) +}; #undef AL asmlinkage long compat_sys_sendmsg(int fd, struct compat_msghdr __user *msg, unsigned flags) @@ -827,7 +824,7 @@ asmlinkage long compat_sys_socketcall(int call, u32 __user *args) compat_ptr(a[4]), compat_ptr(a[5])); break; case SYS_SHUTDOWN: - ret = sys_shutdown(a0,a1); + ret = sys_shutdown(a0, a1); break; case SYS_SETSOCKOPT: ret = compat_sys_setsockopt(a0, a1, a[2], diff --git a/net/socket.c b/net/socket.c index 367d547..b63c051 100644 --- a/net/socket.c +++ b/net/socket.c @@ -124,7 +124,7 @@ static int sock_fasync(int fd, struct file *filp, int on); static ssize_t sock_sendpage(struct file *file, struct page *page, int offset, size_t size, loff_t *ppos, int more); static ssize_t sock_splice_read(struct file *file, loff_t *ppos, - struct pipe_inode_info *pipe, size_t len, + struct pipe_inode_info *pipe, size_t len, unsigned int flags); /* @@ -162,7 +162,7 @@ static const struct net_proto_family *net_families[NPROTO] __read_mostly; * Statistics counters of the socket lists */ -static DEFINE_PER_CPU(int, sockets_in_use) = 0; +static DEFINE_PER_CPU(int, sockets_in_use); /* * Support routines. @@ -309,9 +309,9 @@ static int init_inodecache(void) } static const struct super_operations sockfs_ops = { - .alloc_inode = sock_alloc_inode, - .destroy_inode =sock_destroy_inode, - .statfs = simple_statfs, + .alloc_inode = sock_alloc_inode, + .destroy_inode = sock_destroy_inode, + .statfs = simple_statfs, }; static int sockfs_get_sb(struct file_system_type *fs_type, @@ -411,6 +411,7 @@ int sock_map_fd(struct socket *sock, int flags) return fd; } +EXPORT_SYMBOL(sock_map_fd); static struct socket *sock_from_file(struct file *file, int *err) { @@ -422,7 +423,7 @@ static struct socket *sock_from_file(struct file *file, int *err) } /** - * sockfd_lookup - Go from a file number to its socket slot + * sockfd_lookup - Go from a file number to its socket slot * @fd: file handle * @err: pointer to an error code return * @@ -450,6 +451,7 @@ struct socket *sockfd_lookup(int fd, int *err) fput(file); return sock; } +EXPORT_SYMBOL(sockfd_lookup); static struct socket *sockfd_lookup_light(int fd, int *err, int *fput_needed) { @@ -540,6 +542,7 @@ void sock_release(struct socket *sock) } sock->file = NULL; } +EXPORT_SYMBOL(sock_release); int sock_tx_timestamp(struct msghdr *msg, struct sock *sk, union skb_shared_tx *shtx) @@ -586,6 +589,7 @@ int sock_sendmsg(struct socket *sock, struct msghdr *msg, size_t size) ret = wait_on_sync_kiocb(&iocb); return ret; } +EXPORT_SYMBOL(sock_sendmsg); int kernel_sendmsg(struct socket *sock, struct msghdr *msg, struct kvec *vec, size_t num, size_t size) @@ -604,6 +608,7 @@ int kernel_sendmsg(struct socket *sock, struct msghdr *msg, set_fs(oldfs); return result; } +EXPORT_SYMBOL(kernel_sendmsg); static int ktime2ts(ktime_t kt, struct timespec *ts) { @@ -664,7 +669,6 @@ void __sock_recv_timestamp(struct msghdr *msg, struct sock *sk, put_cmsg(msg, SOL_SOCKET, SCM_TIMESTAMPING, sizeof(ts), &ts); } - EXPORT_SYMBOL_GPL(__sock_recv_timestamp); inline void sock_recv_drops(struct msghdr *msg, struct sock *sk, struct sk_buff *skb) @@ -720,6 +724,7 @@ int sock_recvmsg(struct socket *sock, struct msghdr *msg, ret = wait_on_sync_kiocb(&iocb); return ret; } +EXPORT_SYMBOL(sock_recvmsg); static int sock_recvmsg_nosec(struct socket *sock, struct msghdr *msg, size_t size, int flags) @@ -752,6 +757,7 @@ int kernel_recvmsg(struct socket *sock, struct msghdr *msg, set_fs(oldfs); return result; } +EXPORT_SYMBOL(kernel_recvmsg); static void sock_aio_dtor(struct kiocb *iocb) { @@ -774,7 +780,7 @@ static ssize_t sock_sendpage(struct file *file, struct page *page, } static ssize_t sock_splice_read(struct file *file, loff_t *ppos, - struct pipe_inode_info *pipe, size_t len, + struct pipe_inode_info *pipe, size_t len, unsigned int flags) { struct socket *sock = file->private_data; @@ -887,7 +893,7 @@ static ssize_t sock_aio_write(struct kiocb *iocb, const struct iovec *iov, */ static DEFINE_MUTEX(br_ioctl_mutex); -static int (*br_ioctl_hook) (struct net *, unsigned int cmd, void __user *arg) = NULL; +static int (*br_ioctl_hook) (struct net *, unsigned int cmd, void __user *arg); void brioctl_set(int (*hook) (struct net *, unsigned int, void __user *)) { @@ -895,7 +901,6 @@ void brioctl_set(int (*hook) (struct net *, unsigned int, void __user *)) br_ioctl_hook = hook; mutex_unlock(&br_ioctl_mutex); } - EXPORT_SYMBOL(brioctl_set); static DEFINE_MUTEX(vlan_ioctl_mutex); @@ -907,7 +912,6 @@ void vlan_ioctl_set(int (*hook) (struct net *, void __user *)) vlan_ioctl_hook = hook; mutex_unlock(&vlan_ioctl_mutex); } - EXPORT_SYMBOL(vlan_ioctl_set); static DEFINE_MUTEX(dlci_ioctl_mutex); @@ -919,7 +923,6 @@ void dlci_ioctl_set(int (*hook) (unsigned int, void __user *)) dlci_ioctl_hook = hook; mutex_unlock(&dlci_ioctl_mutex); } - EXPORT_SYMBOL(dlci_ioctl_set); static long sock_do_ioctl(struct net *net, struct socket *sock, @@ -1047,6 +1050,7 @@ out_release: sock = NULL; goto out; } +EXPORT_SYMBOL(sock_create_lite); /* No kernel lock held - perfect */ static unsigned int sock_poll(struct file *file, poll_table *wait) @@ -1147,6 +1151,7 @@ call_kill: rcu_read_unlock(); return 0; } +EXPORT_SYMBOL(sock_wake_async); static int __sock_create(struct net *net, int family, int type, int protocol, struct socket **res, int kern) @@ -1265,11 +1270,13 @@ int sock_create(int family, int type, int protocol, struct socket **res) { return __sock_create(current->nsproxy->net_ns, family, type, protocol, res, 0); } +EXPORT_SYMBOL(sock_create); int sock_create_kern(int family, int type, int protocol, struct socket **res) { return __sock_create(&init_net, family, type, protocol, res, 1); } +EXPORT_SYMBOL(sock_create_kern); SYSCALL_DEFINE3(socket, int, family, int, type, int, protocol) { @@ -1474,7 +1481,8 @@ SYSCALL_DEFINE4(accept4, int, fd, struct sockaddr __user *, upeer_sockaddr, goto out; err = -ENFILE; - if (!(newsock = sock_alloc())) + newsock = sock_alloc(); + if (!newsock) goto out_put; newsock->type = sock->type; @@ -1861,8 +1869,7 @@ SYSCALL_DEFINE3(sendmsg, int, fd, struct msghdr __user *, msg, unsigned, flags) if (MSG_CMSG_COMPAT & flags) { if (get_compat_msghdr(&msg_sys, msg_compat)) return -EFAULT; - } - else if (copy_from_user(&msg_sys, msg, sizeof(struct msghdr))) + } else if (copy_from_user(&msg_sys, msg, sizeof(struct msghdr))) return -EFAULT; sock = sockfd_lookup_light(fd, &err, &fput_needed); @@ -1964,8 +1971,7 @@ static int __sys_recvmsg(struct socket *sock, struct msghdr __user *msg, if (MSG_CMSG_COMPAT & flags) { if (get_compat_msghdr(msg_sys, msg_compat)) return -EFAULT; - } - else if (copy_from_user(msg_sys, msg, sizeof(struct msghdr))) + } else if (copy_from_user(msg_sys, msg, sizeof(struct msghdr))) return -EFAULT; err = -EMSGSIZE; @@ -2191,10 +2197,10 @@ SYSCALL_DEFINE5(recvmmsg, int, fd, struct mmsghdr __user *, mmsg, /* Argument list sizes for sys_socketcall */ #define AL(x) ((x) * sizeof(unsigned long)) static const unsigned char nargs[20] = { - AL(0),AL(3),AL(3),AL(3),AL(2),AL(3), - AL(3),AL(3),AL(4),AL(4),AL(4),AL(6), - AL(6),AL(2),AL(5),AL(5),AL(3),AL(3), - AL(4),AL(5) + AL(0), AL(3), AL(3), AL(3), AL(2), AL(3), + AL(3), AL(3), AL(4), AL(4), AL(4), AL(6), + AL(6), AL(2), AL(5), AL(5), AL(3), AL(3), + AL(4), AL(5) }; #undef AL @@ -2340,6 +2346,7 @@ int sock_register(const struct net_proto_family *ops) printk(KERN_INFO "NET: Registered protocol family %d\n", ops->family); return err; } +EXPORT_SYMBOL(sock_register); /** * sock_unregister - remove a protocol handler @@ -2366,6 +2373,7 @@ void sock_unregister(int family) printk(KERN_INFO "NET: Unregistered protocol family %d\n", family); } +EXPORT_SYMBOL(sock_unregister); static int __init sock_init(void) { @@ -2490,13 +2498,13 @@ static int dev_ifconf(struct net *net, struct compat_ifconf __user *uifc32) ifc.ifc_req = NULL; uifc = compat_alloc_user_space(sizeof(struct ifconf)); } else { - size_t len =((ifc32.ifc_len / sizeof (struct compat_ifreq)) + 1) * - sizeof (struct ifreq); + size_t len = ((ifc32.ifc_len / sizeof(struct compat_ifreq)) + 1) * + sizeof(struct ifreq); uifc = compat_alloc_user_space(sizeof(struct ifconf) + len); ifc.ifc_len = len; ifr = ifc.ifc_req = (void __user *)(uifc + 1); ifr32 = compat_ptr(ifc32.ifcbuf); - for (i = 0; i < ifc32.ifc_len; i += sizeof (struct compat_ifreq)) { + for (i = 0; i < ifc32.ifc_len; i += sizeof(struct compat_ifreq)) { if (copy_in_user(ifr, ifr32, sizeof(struct compat_ifreq))) return -EFAULT; ifr++; @@ -2516,9 +2524,9 @@ static int dev_ifconf(struct net *net, struct compat_ifconf __user *uifc32) ifr = ifc.ifc_req; ifr32 = compat_ptr(ifc32.ifcbuf); for (i = 0, j = 0; - i + sizeof (struct compat_ifreq) <= ifc32.ifc_len && j < ifc.ifc_len; - i += sizeof (struct compat_ifreq), j += sizeof (struct ifreq)) { - if (copy_in_user(ifr32, ifr, sizeof (struct compat_ifreq))) + i + sizeof(struct compat_ifreq) <= ifc32.ifc_len && j < ifc.ifc_len; + i += sizeof(struct compat_ifreq), j += sizeof(struct ifreq)) { + if (copy_in_user(ifr32, ifr, sizeof(struct compat_ifreq))) return -EFAULT; ifr32++; ifr++; @@ -2567,7 +2575,7 @@ static int compat_siocwandev(struct net *net, struct compat_ifreq __user *uifr32 compat_uptr_t uptr32; struct ifreq __user *uifr; - uifr = compat_alloc_user_space(sizeof (*uifr)); + uifr = compat_alloc_user_space(sizeof(*uifr)); if (copy_in_user(uifr, uifr32, sizeof(struct compat_ifreq))) return -EFAULT; @@ -2601,9 +2609,9 @@ static int bond_ioctl(struct net *net, unsigned int cmd, return -EFAULT; old_fs = get_fs(); - set_fs (KERNEL_DS); + set_fs(KERNEL_DS); err = dev_ioctl(net, cmd, &kifr); - set_fs (old_fs); + set_fs(old_fs); return err; case SIOCBONDSLAVEINFOQUERY: @@ -2710,9 +2718,9 @@ static int compat_sioc_ifmap(struct net *net, unsigned int cmd, return -EFAULT; old_fs = get_fs(); - set_fs (KERNEL_DS); + set_fs(KERNEL_DS); err = dev_ioctl(net, cmd, (void __user *)&ifr); - set_fs (old_fs); + set_fs(old_fs); if (cmd == SIOCGIFMAP && !err) { err = copy_to_user(uifr32, &ifr, sizeof(ifr.ifr_name)); @@ -2734,7 +2742,7 @@ static int compat_siocshwtstamp(struct net *net, struct compat_ifreq __user *uif compat_uptr_t uptr32; struct ifreq __user *uifr; - uifr = compat_alloc_user_space(sizeof (*uifr)); + uifr = compat_alloc_user_space(sizeof(*uifr)); if (copy_in_user(uifr, uifr32, sizeof(struct compat_ifreq))) return -EFAULT; @@ -2750,20 +2758,20 @@ static int compat_siocshwtstamp(struct net *net, struct compat_ifreq __user *uif } struct rtentry32 { - u32 rt_pad1; + u32 rt_pad1; struct sockaddr rt_dst; /* target address */ struct sockaddr rt_gateway; /* gateway addr (RTF_GATEWAY) */ struct sockaddr rt_genmask; /* target network mask (IP) */ - unsigned short rt_flags; - short rt_pad2; - u32 rt_pad3; - unsigned char rt_tos; - unsigned char rt_class; - short rt_pad4; - short rt_metric; /* +1 for binary compatibility! */ + unsigned short rt_flags; + short rt_pad2; + u32 rt_pad3; + unsigned char rt_tos; + unsigned char rt_class; + short rt_pad4; + short rt_metric; /* +1 for binary compatibility! */ /* char * */ u32 rt_dev; /* forcing the device at add */ - u32 rt_mtu; /* per route MTU/Window */ - u32 rt_window; /* Window clamping */ + u32 rt_mtu; /* per route MTU/Window */ + u32 rt_window; /* Window clamping */ unsigned short rt_irtt; /* Initial RTT */ }; @@ -2793,29 +2801,29 @@ static int routing_ioctl(struct net *net, struct socket *sock, if (sock && sock->sk && sock->sk->sk_family == AF_INET6) { /* ipv6 */ struct in6_rtmsg32 __user *ur6 = argp; - ret = copy_from_user (&r6.rtmsg_dst, &(ur6->rtmsg_dst), + ret = copy_from_user(&r6.rtmsg_dst, &(ur6->rtmsg_dst), 3 * sizeof(struct in6_addr)); - ret |= __get_user (r6.rtmsg_type, &(ur6->rtmsg_type)); - ret |= __get_user (r6.rtmsg_dst_len, &(ur6->rtmsg_dst_len)); - ret |= __get_user (r6.rtmsg_src_len, &(ur6->rtmsg_src_len)); - ret |= __get_user (r6.rtmsg_metric, &(ur6->rtmsg_metric)); - ret |= __get_user (r6.rtmsg_info, &(ur6->rtmsg_info)); - ret |= __get_user (r6.rtmsg_flags, &(ur6->rtmsg_flags)); - ret |= __get_user (r6.rtmsg_ifindex, &(ur6->rtmsg_ifindex)); + ret |= __get_user(r6.rtmsg_type, &(ur6->rtmsg_type)); + ret |= __get_user(r6.rtmsg_dst_len, &(ur6->rtmsg_dst_len)); + ret |= __get_user(r6.rtmsg_src_len, &(ur6->rtmsg_src_len)); + ret |= __get_user(r6.rtmsg_metric, &(ur6->rtmsg_metric)); + ret |= __get_user(r6.rtmsg_info, &(ur6->rtmsg_info)); + ret |= __get_user(r6.rtmsg_flags, &(ur6->rtmsg_flags)); + ret |= __get_user(r6.rtmsg_ifindex, &(ur6->rtmsg_ifindex)); r = (void *) &r6; } else { /* ipv4 */ struct rtentry32 __user *ur4 = argp; - ret = copy_from_user (&r4.rt_dst, &(ur4->rt_dst), + ret = copy_from_user(&r4.rt_dst, &(ur4->rt_dst), 3 * sizeof(struct sockaddr)); - ret |= __get_user (r4.rt_flags, &(ur4->rt_flags)); - ret |= __get_user (r4.rt_metric, &(ur4->rt_metric)); - ret |= __get_user (r4.rt_mtu, &(ur4->rt_mtu)); - ret |= __get_user (r4.rt_window, &(ur4->rt_window)); - ret |= __get_user (r4.rt_irtt, &(ur4->rt_irtt)); - ret |= __get_user (rtdev, &(ur4->rt_dev)); + ret |= __get_user(r4.rt_flags, &(ur4->rt_flags)); + ret |= __get_user(r4.rt_metric, &(ur4->rt_metric)); + ret |= __get_user(r4.rt_mtu, &(ur4->rt_mtu)); + ret |= __get_user(r4.rt_window, &(ur4->rt_window)); + ret |= __get_user(r4.rt_irtt, &(ur4->rt_irtt)); + ret |= __get_user(rtdev, &(ur4->rt_dev)); if (rtdev) { - ret |= copy_from_user (devname, compat_ptr(rtdev), 15); + ret |= copy_from_user(devname, compat_ptr(rtdev), 15); r4.rt_dev = devname; devname[15] = 0; } else r4.rt_dev = NULL; @@ -2828,9 +2836,9 @@ static int routing_ioctl(struct net *net, struct socket *sock, goto out; } - set_fs (KERNEL_DS); + set_fs(KERNEL_DS); ret = sock_do_ioctl(net, sock, cmd, (unsigned long) r); - set_fs (old_fs); + set_fs(old_fs); out: return ret; @@ -2993,11 +3001,13 @@ int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen) { return sock->ops->bind(sock, addr, addrlen); } +EXPORT_SYMBOL(kernel_bind); int kernel_listen(struct socket *sock, int backlog) { return sock->ops->listen(sock, backlog); } +EXPORT_SYMBOL(kernel_listen); int kernel_accept(struct socket *sock, struct socket **newsock, int flags) { @@ -3022,24 +3032,28 @@ int kernel_accept(struct socket *sock, struct socket **newsock, int flags) done: return err; } +EXPORT_SYMBOL(kernel_accept); int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen, int flags) { return sock->ops->connect(sock, addr, addrlen, flags); } +EXPORT_SYMBOL(kernel_connect); int kernel_getsockname(struct socket *sock, struct sockaddr *addr, int *addrlen) { return sock->ops->getname(sock, addr, addrlen, 0); } +EXPORT_SYMBOL(kernel_getsockname); int kernel_getpeername(struct socket *sock, struct sockaddr *addr, int *addrlen) { return sock->ops->getname(sock, addr, addrlen, 1); } +EXPORT_SYMBOL(kernel_getpeername); int kernel_getsockopt(struct socket *sock, int level, int optname, char *optval, int *optlen) @@ -3056,6 +3070,7 @@ int kernel_getsockopt(struct socket *sock, int level, int optname, set_fs(oldfs); return err; } +EXPORT_SYMBOL(kernel_getsockopt); int kernel_setsockopt(struct socket *sock, int level, int optname, char *optval, unsigned int optlen) @@ -3072,6 +3087,7 @@ int kernel_setsockopt(struct socket *sock, int level, int optname, set_fs(oldfs); return err; } +EXPORT_SYMBOL(kernel_setsockopt); int kernel_sendpage(struct socket *sock, struct page *page, int offset, size_t size, int flags) @@ -3083,6 +3099,7 @@ int kernel_sendpage(struct socket *sock, struct page *page, int offset, return sock_no_sendpage(sock, page, offset, size, flags); } +EXPORT_SYMBOL(kernel_sendpage); int kernel_sock_ioctl(struct socket *sock, int cmd, unsigned long arg) { @@ -3095,33 +3112,11 @@ int kernel_sock_ioctl(struct socket *sock, int cmd, unsigned long arg) return err; } +EXPORT_SYMBOL(kernel_sock_ioctl); int kernel_sock_shutdown(struct socket *sock, enum sock_shutdown_cmd how) { return sock->ops->shutdown(sock, how); } - -EXPORT_SYMBOL(sock_create); -EXPORT_SYMBOL(sock_create_kern); -EXPORT_SYMBOL(sock_create_lite); -EXPORT_SYMBOL(sock_map_fd); -EXPORT_SYMBOL(sock_recvmsg); -EXPORT_SYMBOL(sock_register); -EXPORT_SYMBOL(sock_release); -EXPORT_SYMBOL(sock_sendmsg); -EXPORT_SYMBOL(sock_unregister); -EXPORT_SYMBOL(sock_wake_async); -EXPORT_SYMBOL(sockfd_lookup); -EXPORT_SYMBOL(kernel_sendmsg); -EXPORT_SYMBOL(kernel_recvmsg); -EXPORT_SYMBOL(kernel_bind); -EXPORT_SYMBOL(kernel_listen); -EXPORT_SYMBOL(kernel_accept); -EXPORT_SYMBOL(kernel_connect); -EXPORT_SYMBOL(kernel_getsockname); -EXPORT_SYMBOL(kernel_getpeername); -EXPORT_SYMBOL(kernel_getsockopt); -EXPORT_SYMBOL(kernel_setsockopt); -EXPORT_SYMBOL(kernel_sendpage); -EXPORT_SYMBOL(kernel_sock_ioctl); EXPORT_SYMBOL(kernel_sock_shutdown); + -- 1.7.0.4 --- net/compat.c | 47 ++++++++--------- net/socket.c | 164 ++++++++++++++++++++++++++++------------------------------- 2 files changed, 101 insertions(+), 110 deletions(-) (limited to 'net') diff --git a/net/compat.c b/net/compat.c index 1cf7590..63d260e 100644 --- a/net/compat.c +++ b/net/compat.c @@ -81,7 +81,7 @@ int verify_compat_iovec(struct msghdr *kern_msg, struct iovec *kern_iov, int tot_len; if (kern_msg->msg_namelen) { - if (mode==VERIFY_READ) { + if (mode == VERIFY_READ) { int err = move_addr_to_kernel(kern_msg->msg_name, kern_msg->msg_namelen, kern_address); @@ -354,7 +354,7 @@ static int do_set_attach_filter(struct socket *sock, int level, int optname, static int do_set_sock_timeout(struct socket *sock, int level, int optname, char __user *optval, unsigned int optlen) { - struct compat_timeval __user *up = (struct compat_timeval __user *) optval; + struct compat_timeval __user *up = (struct compat_timeval __user *)optval; struct timeval ktime; mm_segment_t old_fs; int err; @@ -367,7 +367,7 @@ static int do_set_sock_timeout(struct socket *sock, int level, return -EFAULT; old_fs = get_fs(); set_fs(KERNEL_DS); - err = sock_setsockopt(sock, level, optname, (char *) &ktime, sizeof(ktime)); + err = sock_setsockopt(sock, level, optname, (char *)&ktime, sizeof(ktime)); set_fs(old_fs); return err; @@ -389,11 +389,10 @@ asmlinkage long compat_sys_setsockopt(int fd, int level, int optname, char __user *optval, unsigned int optlen) { int err; - struct socket *sock; + struct socket *sock = sockfd_lookup(fd, &err); - if ((sock = sockfd_lookup(fd, &err))!=NULL) - { - err = security_socket_setsockopt(sock,level,optname); + if (sock) { + err = security_socket_setsockopt(sock, level, optname); if (err) { sockfd_put(sock); return err; @@ -453,7 +452,7 @@ static int compat_sock_getsockopt(struct socket *sock, int level, int optname, int compat_sock_get_timestamp(struct sock *sk, struct timeval __user *userstamp) { struct compat_timeval __user *ctv = - (struct compat_timeval __user*) userstamp; + (struct compat_timeval __user *) userstamp; int err = -ENOENT; struct timeval tv; @@ -477,7 +476,7 @@ EXPORT_SYMBOL(compat_sock_get_timestamp); int compat_sock_get_timestampns(struct sock *sk, struct timespec __user *userstamp) { struct compat_timespec __user *ctv = - (struct compat_timespec __user*) userstamp; + (struct compat_timespec __user *) userstamp; int err = -ENOENT; struct timespec ts; @@ -502,12 +501,10 @@ asmlinkage long compat_sys_getsockopt(int fd, int level, int optname, char __user *optval, int __user *optlen) { int err; - struct socket *sock; + struct socket *sock = sockfd_lookup(fd, &err); - if ((sock = sockfd_lookup(fd, &err))!=NULL) - { - err = security_socket_getsockopt(sock, level, - optname); + if (sock) { + err = security_socket_getsockopt(sock, level, optname); if (err) { sockfd_put(sock); return err; @@ -557,7 +554,7 @@ struct compat_group_filter { int compat_mc_setsockopt(struct sock *sock, int level, int optname, char __user *optval, unsigned int optlen, - int (*setsockopt)(struct sock *,int,int,char __user *,unsigned int)) + int (*setsockopt)(struct sock *, int, int, char __user *, unsigned int)) { char __user *koptval = optval; int koptlen = optlen; @@ -640,12 +637,11 @@ int compat_mc_setsockopt(struct sock *sock, int level, int optname, } return setsockopt(sock, level, optname, koptval, koptlen); } - EXPORT_SYMBOL(compat_mc_setsockopt); int compat_mc_getsockopt(struct sock *sock, int level, int optname, char __user *optval, int __user *optlen, - int (*getsockopt)(struct sock *,int,int,char __user *,int __user *)) + int (*getsockopt)(struct sock *, int, int, char __user *, int __user *)) { struct compat_group_filter __user *gf32 = (void *)optval; struct group_filter __user *kgf; @@ -681,7 +677,7 @@ int compat_mc_getsockopt(struct sock *sock, int level, int optname, __put_user(interface, &kgf->gf_interface) || __put_user(fmode, &kgf->gf_fmode) || __put_user(numsrc, &kgf->gf_numsrc) || - copy_in_user(&kgf->gf_group,&gf32->gf_group,sizeof(kgf->gf_group))) + copy_in_user(&kgf->gf_group, &gf32->gf_group, sizeof(kgf->gf_group))) return -EFAULT; err = getsockopt(sock, level, optname, (char __user *)kgf, koptlen); @@ -714,21 +710,22 @@ int compat_mc_getsockopt(struct sock *sock, int level, int optname, copylen = numsrc * sizeof(gf32->gf_slist[0]); if (copylen > klen) copylen = klen; - if (copy_in_user(gf32->gf_slist, kgf->gf_slist, copylen)) + if (copy_in_user(gf32->gf_slist, kgf->gf_slist, copylen)) return -EFAULT; } return err; } - EXPORT_SYMBOL(compat_mc_getsockopt); /* Argument list sizes for compat_sys_socketcall */ #define AL(x) ((x) * sizeof(u32)) -static unsigned char nas[20]={AL(0),AL(3),AL(3),AL(3),AL(2),AL(3), - AL(3),AL(3),AL(4),AL(4),AL(4),AL(6), - AL(6),AL(2),AL(5),AL(5),AL(3),AL(3), - AL(4),AL(5)}; +static unsigned char nas[20] = { + AL(0), AL(3), AL(3), AL(3), AL(2), AL(3), + AL(3), AL(3), AL(4), AL(4), AL(4), AL(6), + AL(6), AL(2), AL(5), AL(5), AL(3), AL(3), + AL(4), AL(5) +}; #undef AL asmlinkage long compat_sys_sendmsg(int fd, struct compat_msghdr __user *msg, unsigned flags) @@ -827,7 +824,7 @@ asmlinkage long compat_sys_socketcall(int call, u32 __user *args) compat_ptr(a[4]), compat_ptr(a[5])); break; case SYS_SHUTDOWN: - ret = sys_shutdown(a0,a1); + ret = sys_shutdown(a0, a1); break; case SYS_SETSOCKOPT: ret = compat_sys_setsockopt(a0, a1, a[2], diff --git a/net/socket.c b/net/socket.c index 367d547..acfa173 100644 --- a/net/socket.c +++ b/net/socket.c @@ -124,7 +124,7 @@ static int sock_fasync(int fd, struct file *filp, int on); static ssize_t sock_sendpage(struct file *file, struct page *page, int offset, size_t size, loff_t *ppos, int more); static ssize_t sock_splice_read(struct file *file, loff_t *ppos, - struct pipe_inode_info *pipe, size_t len, + struct pipe_inode_info *pipe, size_t len, unsigned int flags); /* @@ -162,7 +162,7 @@ static const struct net_proto_family *net_families[NPROTO] __read_mostly; * Statistics counters of the socket lists */ -static DEFINE_PER_CPU(int, sockets_in_use) = 0; +static DEFINE_PER_CPU(int, sockets_in_use); /* * Support routines. @@ -309,9 +309,9 @@ static int init_inodecache(void) } static const struct super_operations sockfs_ops = { - .alloc_inode = sock_alloc_inode, - .destroy_inode =sock_destroy_inode, - .statfs = simple_statfs, + .alloc_inode = sock_alloc_inode, + .destroy_inode = sock_destroy_inode, + .statfs = simple_statfs, }; static int sockfs_get_sb(struct file_system_type *fs_type, @@ -411,6 +411,7 @@ int sock_map_fd(struct socket *sock, int flags) return fd; } +EXPORT_SYMBOL(sock_map_fd); static struct socket *sock_from_file(struct file *file, int *err) { @@ -422,7 +423,7 @@ static struct socket *sock_from_file(struct file *file, int *err) } /** - * sockfd_lookup - Go from a file number to its socket slot + * sockfd_lookup - Go from a file number to its socket slot * @fd: file handle * @err: pointer to an error code return * @@ -450,6 +451,7 @@ struct socket *sockfd_lookup(int fd, int *err) fput(file); return sock; } +EXPORT_SYMBOL(sockfd_lookup); static struct socket *sockfd_lookup_light(int fd, int *err, int *fput_needed) { @@ -540,6 +542,7 @@ void sock_release(struct socket *sock) } sock->file = NULL; } +EXPORT_SYMBOL(sock_release); int sock_tx_timestamp(struct msghdr *msg, struct sock *sk, union skb_shared_tx *shtx) @@ -586,6 +589,7 @@ int sock_sendmsg(struct socket *sock, struct msghdr *msg, size_t size) ret = wait_on_sync_kiocb(&iocb); return ret; } +EXPORT_SYMBOL(sock_sendmsg); int kernel_sendmsg(struct socket *sock, struct msghdr *msg, struct kvec *vec, size_t num, size_t size) @@ -604,6 +608,7 @@ int kernel_sendmsg(struct socket *sock, struct msghdr *msg, set_fs(oldfs); return result; } +EXPORT_SYMBOL(kernel_sendmsg); static int ktime2ts(ktime_t kt, struct timespec *ts) { @@ -664,7 +669,6 @@ void __sock_recv_timestamp(struct msghdr *msg, struct sock *sk, put_cmsg(msg, SOL_SOCKET, SCM_TIMESTAMPING, sizeof(ts), &ts); } - EXPORT_SYMBOL_GPL(__sock_recv_timestamp); inline void sock_recv_drops(struct msghdr *msg, struct sock *sk, struct sk_buff *skb) @@ -720,6 +724,7 @@ int sock_recvmsg(struct socket *sock, struct msghdr *msg, ret = wait_on_sync_kiocb(&iocb); return ret; } +EXPORT_SYMBOL(sock_recvmsg); static int sock_recvmsg_nosec(struct socket *sock, struct msghdr *msg, size_t size, int flags) @@ -752,6 +757,7 @@ int kernel_recvmsg(struct socket *sock, struct msghdr *msg, set_fs(oldfs); return result; } +EXPORT_SYMBOL(kernel_recvmsg); static void sock_aio_dtor(struct kiocb *iocb) { @@ -774,7 +780,7 @@ static ssize_t sock_sendpage(struct file *file, struct page *page, } static ssize_t sock_splice_read(struct file *file, loff_t *ppos, - struct pipe_inode_info *pipe, size_t len, + struct pipe_inode_info *pipe, size_t len, unsigned int flags) { struct socket *sock = file->private_data; @@ -887,7 +893,7 @@ static ssize_t sock_aio_write(struct kiocb *iocb, const struct iovec *iov, */ static DEFINE_MUTEX(br_ioctl_mutex); -static int (*br_ioctl_hook) (struct net *, unsigned int cmd, void __user *arg) = NULL; +static int (*br_ioctl_hook) (struct net *, unsigned int cmd, void __user *arg); void brioctl_set(int (*hook) (struct net *, unsigned int, void __user *)) { @@ -895,7 +901,6 @@ void brioctl_set(int (*hook) (struct net *, unsigned int, void __user *)) br_ioctl_hook = hook; mutex_unlock(&br_ioctl_mutex); } - EXPORT_SYMBOL(brioctl_set); static DEFINE_MUTEX(vlan_ioctl_mutex); @@ -907,7 +912,6 @@ void vlan_ioctl_set(int (*hook) (struct net *, void __user *)) vlan_ioctl_hook = hook; mutex_unlock(&vlan_ioctl_mutex); } - EXPORT_SYMBOL(vlan_ioctl_set); static DEFINE_MUTEX(dlci_ioctl_mutex); @@ -919,7 +923,6 @@ void dlci_ioctl_set(int (*hook) (unsigned int, void __user *)) dlci_ioctl_hook = hook; mutex_unlock(&dlci_ioctl_mutex); } - EXPORT_SYMBOL(dlci_ioctl_set); static long sock_do_ioctl(struct net *net, struct socket *sock, @@ -1047,6 +1050,7 @@ out_release: sock = NULL; goto out; } +EXPORT_SYMBOL(sock_create_lite); /* No kernel lock held - perfect */ static unsigned int sock_poll(struct file *file, poll_table *wait) @@ -1147,6 +1151,7 @@ call_kill: rcu_read_unlock(); return 0; } +EXPORT_SYMBOL(sock_wake_async); static int __sock_create(struct net *net, int family, int type, int protocol, struct socket **res, int kern) @@ -1265,11 +1270,13 @@ int sock_create(int family, int type, int protocol, struct socket **res) { return __sock_create(current->nsproxy->net_ns, family, type, protocol, res, 0); } +EXPORT_SYMBOL(sock_create); int sock_create_kern(int family, int type, int protocol, struct socket **res) { return __sock_create(&init_net, family, type, protocol, res, 1); } +EXPORT_SYMBOL(sock_create_kern); SYSCALL_DEFINE3(socket, int, family, int, type, int, protocol) { @@ -1474,7 +1481,8 @@ SYSCALL_DEFINE4(accept4, int, fd, struct sockaddr __user *, upeer_sockaddr, goto out; err = -ENFILE; - if (!(newsock = sock_alloc())) + newsock = sock_alloc(); + if (!newsock) goto out_put; newsock->type = sock->type; @@ -1861,8 +1869,7 @@ SYSCALL_DEFINE3(sendmsg, int, fd, struct msghdr __user *, msg, unsigned, flags) if (MSG_CMSG_COMPAT & flags) { if (get_compat_msghdr(&msg_sys, msg_compat)) return -EFAULT; - } - else if (copy_from_user(&msg_sys, msg, sizeof(struct msghdr))) + } else if (copy_from_user(&msg_sys, msg, sizeof(struct msghdr))) return -EFAULT; sock = sockfd_lookup_light(fd, &err, &fput_needed); @@ -1964,8 +1971,7 @@ static int __sys_recvmsg(struct socket *sock, struct msghdr __user *msg, if (MSG_CMSG_COMPAT & flags) { if (get_compat_msghdr(msg_sys, msg_compat)) return -EFAULT; - } - else if (copy_from_user(msg_sys, msg, sizeof(struct msghdr))) + } else if (copy_from_user(msg_sys, msg, sizeof(struct msghdr))) return -EFAULT; err = -EMSGSIZE; @@ -2191,10 +2197,10 @@ SYSCALL_DEFINE5(recvmmsg, int, fd, struct mmsghdr __user *, mmsg, /* Argument list sizes for sys_socketcall */ #define AL(x) ((x) * sizeof(unsigned long)) static const unsigned char nargs[20] = { - AL(0),AL(3),AL(3),AL(3),AL(2),AL(3), - AL(3),AL(3),AL(4),AL(4),AL(4),AL(6), - AL(6),AL(2),AL(5),AL(5),AL(3),AL(3), - AL(4),AL(5) + AL(0), AL(3), AL(3), AL(3), AL(2), AL(3), + AL(3), AL(3), AL(4), AL(4), AL(4), AL(6), + AL(6), AL(2), AL(5), AL(5), AL(3), AL(3), + AL(4), AL(5) }; #undef AL @@ -2340,6 +2346,7 @@ int sock_register(const struct net_proto_family *ops) printk(KERN_INFO "NET: Registered protocol family %d\n", ops->family); return err; } +EXPORT_SYMBOL(sock_register); /** * sock_unregister - remove a protocol handler @@ -2366,6 +2373,7 @@ void sock_unregister(int family) printk(KERN_INFO "NET: Unregistered protocol family %d\n", family); } +EXPORT_SYMBOL(sock_unregister); static int __init sock_init(void) { @@ -2490,13 +2498,13 @@ static int dev_ifconf(struct net *net, struct compat_ifconf __user *uifc32) ifc.ifc_req = NULL; uifc = compat_alloc_user_space(sizeof(struct ifconf)); } else { - size_t len =((ifc32.ifc_len / sizeof (struct compat_ifreq)) + 1) * - sizeof (struct ifreq); + size_t len = ((ifc32.ifc_len / sizeof(struct compat_ifreq)) + 1) * + sizeof(struct ifreq); uifc = compat_alloc_user_space(sizeof(struct ifconf) + len); ifc.ifc_len = len; ifr = ifc.ifc_req = (void __user *)(uifc + 1); ifr32 = compat_ptr(ifc32.ifcbuf); - for (i = 0; i < ifc32.ifc_len; i += sizeof (struct compat_ifreq)) { + for (i = 0; i < ifc32.ifc_len; i += sizeof(struct compat_ifreq)) { if (copy_in_user(ifr, ifr32, sizeof(struct compat_ifreq))) return -EFAULT; ifr++; @@ -2516,9 +2524,9 @@ static int dev_ifconf(struct net *net, struct compat_ifconf __user *uifc32) ifr = ifc.ifc_req; ifr32 = compat_ptr(ifc32.ifcbuf); for (i = 0, j = 0; - i + sizeof (struct compat_ifreq) <= ifc32.ifc_len && j < ifc.ifc_len; - i += sizeof (struct compat_ifreq), j += sizeof (struct ifreq)) { - if (copy_in_user(ifr32, ifr, sizeof (struct compat_ifreq))) + i + sizeof(struct compat_ifreq) <= ifc32.ifc_len && j < ifc.ifc_len; + i += sizeof(struct compat_ifreq), j += sizeof(struct ifreq)) { + if (copy_in_user(ifr32, ifr, sizeof(struct compat_ifreq))) return -EFAULT; ifr32++; ifr++; @@ -2567,7 +2575,7 @@ static int compat_siocwandev(struct net *net, struct compat_ifreq __user *uifr32 compat_uptr_t uptr32; struct ifreq __user *uifr; - uifr = compat_alloc_user_space(sizeof (*uifr)); + uifr = compat_alloc_user_space(sizeof(*uifr)); if (copy_in_user(uifr, uifr32, sizeof(struct compat_ifreq))) return -EFAULT; @@ -2601,9 +2609,9 @@ static int bond_ioctl(struct net *net, unsigned int cmd, return -EFAULT; old_fs = get_fs(); - set_fs (KERNEL_DS); + set_fs(KERNEL_DS); err = dev_ioctl(net, cmd, &kifr); - set_fs (old_fs); + set_fs(old_fs); return err; case SIOCBONDSLAVEINFOQUERY: @@ -2710,9 +2718,9 @@ static int compat_sioc_ifmap(struct net *net, unsigned int cmd, return -EFAULT; old_fs = get_fs(); - set_fs (KERNEL_DS); + set_fs(KERNEL_DS); err = dev_ioctl(net, cmd, (void __user *)&ifr); - set_fs (old_fs); + set_fs(old_fs); if (cmd == SIOCGIFMAP && !err) { err = copy_to_user(uifr32, &ifr, sizeof(ifr.ifr_name)); @@ -2734,7 +2742,7 @@ static int compat_siocshwtstamp(struct net *net, struct compat_ifreq __user *uif compat_uptr_t uptr32; struct ifreq __user *uifr; - uifr = compat_alloc_user_space(sizeof (*uifr)); + uifr = compat_alloc_user_space(sizeof(*uifr)); if (copy_in_user(uifr, uifr32, sizeof(struct compat_ifreq))) return -EFAULT; @@ -2750,20 +2758,20 @@ static int compat_siocshwtstamp(struct net *net, struct compat_ifreq __user *uif } struct rtentry32 { - u32 rt_pad1; + u32 rt_pad1; struct sockaddr rt_dst; /* target address */ struct sockaddr rt_gateway; /* gateway addr (RTF_GATEWAY) */ struct sockaddr rt_genmask; /* target network mask (IP) */ - unsigned short rt_flags; - short rt_pad2; - u32 rt_pad3; - unsigned char rt_tos; - unsigned char rt_class; - short rt_pad4; - short rt_metric; /* +1 for binary compatibility! */ + unsigned short rt_flags; + short rt_pad2; + u32 rt_pad3; + unsigned char rt_tos; + unsigned char rt_class; + short rt_pad4; + short rt_metric; /* +1 for binary compatibility! */ /* char * */ u32 rt_dev; /* forcing the device at add */ - u32 rt_mtu; /* per route MTU/Window */ - u32 rt_window; /* Window clamping */ + u32 rt_mtu; /* per route MTU/Window */ + u32 rt_window; /* Window clamping */ unsigned short rt_irtt; /* Initial RTT */ }; @@ -2793,29 +2801,29 @@ static int routing_ioctl(struct net *net, struct socket *sock, if (sock && sock->sk && sock->sk->sk_family == AF_INET6) { /* ipv6 */ struct in6_rtmsg32 __user *ur6 = argp; - ret = copy_from_user (&r6.rtmsg_dst, &(ur6->rtmsg_dst), + ret = copy_from_user(&r6.rtmsg_dst, &(ur6->rtmsg_dst), 3 * sizeof(struct in6_addr)); - ret |= __get_user (r6.rtmsg_type, &(ur6->rtmsg_type)); - ret |= __get_user (r6.rtmsg_dst_len, &(ur6->rtmsg_dst_len)); - ret |= __get_user (r6.rtmsg_src_len, &(ur6->rtmsg_src_len)); - ret |= __get_user (r6.rtmsg_metric, &(ur6->rtmsg_metric)); - ret |= __get_user (r6.rtmsg_info, &(ur6->rtmsg_info)); - ret |= __get_user (r6.rtmsg_flags, &(ur6->rtmsg_flags)); - ret |= __get_user (r6.rtmsg_ifindex, &(ur6->rtmsg_ifindex)); + ret |= __get_user(r6.rtmsg_type, &(ur6->rtmsg_type)); + ret |= __get_user(r6.rtmsg_dst_len, &(ur6->rtmsg_dst_len)); + ret |= __get_user(r6.rtmsg_src_len, &(ur6->rtmsg_src_len)); + ret |= __get_user(r6.rtmsg_metric, &(ur6->rtmsg_metric)); + ret |= __get_user(r6.rtmsg_info, &(ur6->rtmsg_info)); + ret |= __get_user(r6.rtmsg_flags, &(ur6->rtmsg_flags)); + ret |= __get_user(r6.rtmsg_ifindex, &(ur6->rtmsg_ifindex)); r = (void *) &r6; } else { /* ipv4 */ struct rtentry32 __user *ur4 = argp; - ret = copy_from_user (&r4.rt_dst, &(ur4->rt_dst), + ret = copy_from_user(&r4.rt_dst, &(ur4->rt_dst), 3 * sizeof(struct sockaddr)); - ret |= __get_user (r4.rt_flags, &(ur4->rt_flags)); - ret |= __get_user (r4.rt_metric, &(ur4->rt_metric)); - ret |= __get_user (r4.rt_mtu, &(ur4->rt_mtu)); - ret |= __get_user (r4.rt_window, &(ur4->rt_window)); - ret |= __get_user (r4.rt_irtt, &(ur4->rt_irtt)); - ret |= __get_user (rtdev, &(ur4->rt_dev)); + ret |= __get_user(r4.rt_flags, &(ur4->rt_flags)); + ret |= __get_user(r4.rt_metric, &(ur4->rt_metric)); + ret |= __get_user(r4.rt_mtu, &(ur4->rt_mtu)); + ret |= __get_user(r4.rt_window, &(ur4->rt_window)); + ret |= __get_user(r4.rt_irtt, &(ur4->rt_irtt)); + ret |= __get_user(rtdev, &(ur4->rt_dev)); if (rtdev) { - ret |= copy_from_user (devname, compat_ptr(rtdev), 15); + ret |= copy_from_user(devname, compat_ptr(rtdev), 15); r4.rt_dev = devname; devname[15] = 0; } else r4.rt_dev = NULL; @@ -2828,9 +2836,9 @@ static int routing_ioctl(struct net *net, struct socket *sock, goto out; } - set_fs (KERNEL_DS); + set_fs(KERNEL_DS); ret = sock_do_ioctl(net, sock, cmd, (unsigned long) r); - set_fs (old_fs); + set_fs(old_fs); out: return ret; @@ -2993,11 +3001,13 @@ int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen) { return sock->ops->bind(sock, addr, addrlen); } +EXPORT_SYMBOL(kernel_bind); int kernel_listen(struct socket *sock, int backlog) { return sock->ops->listen(sock, backlog); } +EXPORT_SYMBOL(kernel_listen); int kernel_accept(struct socket *sock, struct socket **newsock, int flags) { @@ -3022,24 +3032,28 @@ int kernel_accept(struct socket *sock, struct socket **newsock, int flags) done: return err; } +EXPORT_SYMBOL(kernel_accept); int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen, int flags) { return sock->ops->connect(sock, addr, addrlen, flags); } +EXPORT_SYMBOL(kernel_connect); int kernel_getsockname(struct socket *sock, struct sockaddr *addr, int *addrlen) { return sock->ops->getname(sock, addr, addrlen, 0); } +EXPORT_SYMBOL(kernel_getsockname); int kernel_getpeername(struct socket *sock, struct sockaddr *addr, int *addrlen) { return sock->ops->getname(sock, addr, addrlen, 1); } +EXPORT_SYMBOL(kernel_getpeername); int kernel_getsockopt(struct socket *sock, int level, int optname, char *optval, int *optlen) @@ -3056,6 +3070,7 @@ int kernel_getsockopt(struct socket *sock, int level, int optname, set_fs(oldfs); return err; } +EXPORT_SYMBOL(kernel_getsockopt); int kernel_setsockopt(struct socket *sock, int level, int optname, char *optval, unsigned int optlen) @@ -3072,6 +3087,7 @@ int kernel_setsockopt(struct socket *sock, int level, int optname, set_fs(oldfs); return err; } +EXPORT_SYMBOL(kernel_setsockopt); int kernel_sendpage(struct socket *sock, struct page *page, int offset, size_t size, int flags) @@ -3083,6 +3099,7 @@ int kernel_sendpage(struct socket *sock, struct page *page, int offset, return sock_no_sendpage(sock, page, offset, size, flags); } +EXPORT_SYMBOL(kernel_sendpage); int kernel_sock_ioctl(struct socket *sock, int cmd, unsigned long arg) { @@ -3095,33 +3112,10 @@ int kernel_sock_ioctl(struct socket *sock, int cmd, unsigned long arg) return err; } +EXPORT_SYMBOL(kernel_sock_ioctl); int kernel_sock_shutdown(struct socket *sock, enum sock_shutdown_cmd how) { return sock->ops->shutdown(sock, how); } - -EXPORT_SYMBOL(sock_create); -EXPORT_SYMBOL(sock_create_kern); -EXPORT_SYMBOL(sock_create_lite); -EXPORT_SYMBOL(sock_map_fd); -EXPORT_SYMBOL(sock_recvmsg); -EXPORT_SYMBOL(sock_register); -EXPORT_SYMBOL(sock_release); -EXPORT_SYMBOL(sock_sendmsg); -EXPORT_SYMBOL(sock_unregister); -EXPORT_SYMBOL(sock_wake_async); -EXPORT_SYMBOL(sockfd_lookup); -EXPORT_SYMBOL(kernel_sendmsg); -EXPORT_SYMBOL(kernel_recvmsg); -EXPORT_SYMBOL(kernel_bind); -EXPORT_SYMBOL(kernel_listen); -EXPORT_SYMBOL(kernel_accept); -EXPORT_SYMBOL(kernel_connect); -EXPORT_SYMBOL(kernel_getsockname); -EXPORT_SYMBOL(kernel_getpeername); -EXPORT_SYMBOL(kernel_getsockopt); -EXPORT_SYMBOL(kernel_setsockopt); -EXPORT_SYMBOL(kernel_sendpage); -EXPORT_SYMBOL(kernel_sock_ioctl); EXPORT_SYMBOL(kernel_sock_shutdown); -- cgit v1.1 From e12f8e29a8526172b7715881503bae636d60bdd8 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 4 Jun 2010 13:31:29 +0200 Subject: netfilter: vmalloc_node cleanup Using vmalloc_node(size, numa_node_id()) for temporary storage is not needed. vmalloc(size) is more respectful of user NUMA policy. Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/ipv4/netfilter/arp_tables.c | 7 +++---- net/ipv4/netfilter/ip_tables.c | 4 ++-- net/ipv6/netfilter/ip6_tables.c | 7 +++---- 3 files changed, 8 insertions(+), 10 deletions(-) (limited to 'net') diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c index 1ac01b1..16c0ba0 100644 --- a/net/ipv4/netfilter/arp_tables.c +++ b/net/ipv4/netfilter/arp_tables.c @@ -758,7 +758,7 @@ static struct xt_counters *alloc_counters(const struct xt_table *table) * about). */ countersize = sizeof(struct xt_counters) * private->number; - counters = vmalloc_node(countersize, numa_node_id()); + counters = vmalloc(countersize); if (counters == NULL) return ERR_PTR(-ENOMEM); @@ -1005,8 +1005,7 @@ static int __do_replace(struct net *net, const char *name, struct arpt_entry *iter; ret = 0; - counters = vmalloc_node(num_counters * sizeof(struct xt_counters), - numa_node_id()); + counters = vmalloc(num_counters * sizeof(struct xt_counters)); if (!counters) { ret = -ENOMEM; goto out; @@ -1159,7 +1158,7 @@ static int do_add_counters(struct net *net, const void __user *user, if (len != size + num_counters * sizeof(struct xt_counters)) return -EINVAL; - paddc = vmalloc_node(len - size, numa_node_id()); + paddc = vmalloc(len - size); if (!paddc) return -ENOMEM; diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c index 63958f3..7c0b8ad6 100644 --- a/net/ipv4/netfilter/ip_tables.c +++ b/net/ipv4/netfilter/ip_tables.c @@ -928,7 +928,7 @@ static struct xt_counters *alloc_counters(const struct xt_table *table) (other than comefrom, which userspace doesn't care about). */ countersize = sizeof(struct xt_counters) * private->number; - counters = vmalloc_node(countersize, numa_node_id()); + counters = vmalloc(countersize); if (counters == NULL) return ERR_PTR(-ENOMEM); @@ -1352,7 +1352,7 @@ do_add_counters(struct net *net, const void __user *user, if (len != size + num_counters * sizeof(struct xt_counters)) return -EINVAL; - paddc = vmalloc_node(len - size, numa_node_id()); + paddc = vmalloc(len - size); if (!paddc) return -ENOMEM; diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c index 6f517bd..82945ef 100644 --- a/net/ipv6/netfilter/ip6_tables.c +++ b/net/ipv6/netfilter/ip6_tables.c @@ -943,7 +943,7 @@ static struct xt_counters *alloc_counters(const struct xt_table *table) (other than comefrom, which userspace doesn't care about). */ countersize = sizeof(struct xt_counters) * private->number; - counters = vmalloc_node(countersize, numa_node_id()); + counters = vmalloc(countersize); if (counters == NULL) return ERR_PTR(-ENOMEM); @@ -1213,8 +1213,7 @@ __do_replace(struct net *net, const char *name, unsigned int valid_hooks, struct ip6t_entry *iter; ret = 0; - counters = vmalloc_node(num_counters * sizeof(struct xt_counters), - numa_node_id()); + counters = vmalloc(num_counters * sizeof(struct xt_counters)); if (!counters) { ret = -ENOMEM; goto out; @@ -1368,7 +1367,7 @@ do_add_counters(struct net *net, const void __user *user, unsigned int len, if (len != size + num_counters * sizeof(struct xt_counters)) return -EINVAL; - paddc = vmalloc_node(len - size, numa_node_id()); + paddc = vmalloc(len - size); if (!paddc) return -ENOMEM; -- cgit v1.1 From 26b36cfefaf2be98b225e3c1a399edb0daf52ddd Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Fri, 4 Jun 2010 14:25:44 -0400 Subject: mac80211: make ARP filtering depend on CONFIG_INET Signed-off-by: John W. Linville --- net/mac80211/main.c | 6 ++++++ net/mac80211/mlme.c | 2 ++ 2 files changed, 8 insertions(+) (limited to 'net') diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 045ead9..5706156 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -329,6 +329,7 @@ static void ieee80211_recalc_smps_work(struct work_struct *work) mutex_unlock(&local->iflist_mtx); } +#ifdef CONFIG_INET int ieee80211_set_arp_filter(struct ieee80211_sub_if_data *sdata) { struct in_device *idev; @@ -380,6 +381,7 @@ static int ieee80211_ifa_changed(struct notifier_block *nb, return NOTIFY_DONE; } +#endif struct ieee80211_hw *ieee80211_alloc_hw(size_t priv_data_len, const struct ieee80211_ops *ops) @@ -669,10 +671,12 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) goto fail_pm_qos; } +#ifdef CONFIG_INET local->ifa_notifier.notifier_call = ieee80211_ifa_changed; result = register_inetaddr_notifier(&local->ifa_notifier); if (result) goto fail_ifa; +#endif return 0; @@ -707,7 +711,9 @@ void ieee80211_unregister_hw(struct ieee80211_hw *hw) pm_qos_remove_notifier(PM_QOS_NETWORK_LATENCY, &local->network_latency_notifier); +#ifdef CONFIG_INET unregister_inetaddr_notifier(&local->ifa_notifier); +#endif rtnl_lock(); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 7e72013..3623bb7 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2078,6 +2078,7 @@ static enum work_done_result ieee80211_assoc_done(struct ieee80211_work *wk, cfg80211_send_assoc_timeout(wk->sdata->dev, wk->filter_ta); return WORK_DONE_DESTROY; +#ifdef CONFIG_INET } else { mutex_unlock(&wk->sdata->u.mgd.mtx); @@ -2088,6 +2089,7 @@ static enum work_done_result ieee80211_assoc_done(struct ieee80211_work *wk, rtnl_lock(); ieee80211_set_arp_filter(wk->sdata); rtnl_unlock(); +#endif } } -- cgit v1.1 From 38a6cc7538d3c44b76f9dcea607a171adcc0208e Mon Sep 17 00:00:00 2001 From: Sujith Date: Wed, 19 May 2010 11:32:30 +0530 Subject: mac80211: Remove deprecated sta_notify commands STA_NOTIFY_ADD and STA_NOTIFY_REMOVE have no users anymore, and station addition/removal are indicated to drivers using sta_add() and sta_remove(), which can sleep. Signed-off-by: Sujith Acked-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/driver-ops.h | 6 ------ 1 file changed, 6 deletions(-) (limited to 'net') diff --git a/net/mac80211/driver-ops.h b/net/mac80211/driver-ops.h index 978850e..d1139e4 100644 --- a/net/mac80211/driver-ops.h +++ b/net/mac80211/driver-ops.h @@ -269,9 +269,6 @@ static inline int drv_sta_add(struct ieee80211_local *local, if (local->ops->sta_add) ret = local->ops->sta_add(&local->hw, &sdata->vif, sta); - else if (local->ops->sta_notify) - local->ops->sta_notify(&local->hw, &sdata->vif, - STA_NOTIFY_ADD, sta); trace_drv_sta_add(local, sdata, sta, ret); @@ -286,9 +283,6 @@ static inline void drv_sta_remove(struct ieee80211_local *local, if (local->ops->sta_remove) local->ops->sta_remove(&local->hw, &sdata->vif, sta); - else if (local->ops->sta_notify) - local->ops->sta_notify(&local->hw, &sdata->vif, - STA_NOTIFY_REMOVE, sta); trace_drv_sta_remove(local, sdata, sta); } -- cgit v1.1 From 2a1d4bd46047efff513600d7ff422bc344f540a6 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 3 Jun 2010 00:43:12 +0000 Subject: syncookies: make v4/v6 synflood warning behaviour the same both syn_flood_warning functions print a message, but ipv4 version only prints a warning if CONFIG_SYN_COOKIES=y. Make the v4 one behave like the v6 one. Signed-off-by: Florian Westphal Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 24 +++++++++++++----------- 1 file changed, 13 insertions(+), 11 deletions(-) (limited to 'net') diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 202cf09..a13f881 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -793,19 +793,20 @@ static void tcp_v4_reqsk_destructor(struct request_sock *req) kfree(inet_rsk(req)->opt); } -#ifdef CONFIG_SYN_COOKIES -static void syn_flood_warning(struct sk_buff *skb) +static void syn_flood_warning(const struct sk_buff *skb) { - static unsigned long warntime; + const char *msg; - if (time_after(jiffies, (warntime + HZ * 60))) { - warntime = jiffies; - printk(KERN_INFO - "possible SYN flooding on port %d. Sending cookies.\n", - ntohs(tcp_hdr(skb)->dest)); - } -} +#ifdef CONFIG_SYN_COOKIES + if (sysctl_tcp_syncookies) + msg = "Sending cookies"; + else #endif + msg = "Dropping request"; + + pr_info("TCP: Possible SYN flooding on port %d. %s.\n", + ntohs(tcp_hdr(skb)->dest), msg); +} /* * Save and compile IPv4 options into the request_sock if needed. @@ -1243,6 +1244,8 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) * evidently real one. */ if (inet_csk_reqsk_queue_is_full(sk) && !isn) { + if (net_ratelimit()) + syn_flood_warning(skb); #ifdef CONFIG_SYN_COOKIES if (sysctl_tcp_syncookies) { want_cookie = 1; @@ -1328,7 +1331,6 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) if (want_cookie) { #ifdef CONFIG_SYN_COOKIES - syn_flood_warning(skb); req->cookie_ts = tmp_opt.tstamp_ok; #endif isn = cookie_v4_init_sequence(sk, skb, &req->mss); -- cgit v1.1 From af9b4738574b46025de7ccbe75c7b24fd8914379 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 3 Jun 2010 00:43:44 +0000 Subject: syncookies: avoid unneeded tcp header flag double check caller: if (!th->rst && !th->syn && th->ack) callee: if (!th->ack) make the caller only check for !syn (common for 3whs), and move the !rst / ack test to the callee. Signed-off-by: Florian Westphal Signed-off-by: David S. Miller --- net/ipv4/syncookies.c | 2 +- net/ipv4/tcp_ipv4.c | 2 +- net/ipv6/syncookies.c | 2 +- net/ipv6/tcp_ipv6.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c index 5c24db4..c9dac86 100644 --- a/net/ipv4/syncookies.c +++ b/net/ipv4/syncookies.c @@ -266,7 +266,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb, struct rtable *rt; __u8 rcv_wscale; - if (!sysctl_tcp_syncookies || !th->ack) + if (!sysctl_tcp_syncookies || !th->ack || th->rst) goto out; if (tcp_synq_no_recent_overflow(sk) || diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index a13f881..6558dfd 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1506,7 +1506,7 @@ static struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb) } #ifdef CONFIG_SYN_COOKIES - if (!th->rst && !th->syn && th->ack) + if (!th->syn) sk = cookie_v4_check(sk, skb, &(IPCB(skb)->opt)); #endif return sk; diff --git a/net/ipv6/syncookies.c b/net/ipv6/syncookies.c index 1238370..9fcb3ec 100644 --- a/net/ipv6/syncookies.c +++ b/net/ipv6/syncookies.c @@ -174,7 +174,7 @@ struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb) struct dst_entry *dst; __u8 rcv_wscale; - if (!sysctl_tcp_syncookies || !th->ack) + if (!sysctl_tcp_syncookies || !th->ack || th->rst) goto out; if (tcp_synq_no_recent_overflow(sk) || diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index e487080..5887141 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1157,7 +1157,7 @@ static struct sock *tcp_v6_hnd_req(struct sock *sk,struct sk_buff *skb) } #ifdef CONFIG_SYN_COOKIES - if (!th->rst && !th->syn && th->ack) + if (!th->syn) sk = cookie_v6_check(sk, skb); #endif return sk; -- cgit v1.1 From 5918e2fb9035b35164002c3a268feaf70b7c04d5 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 3 Jun 2010 00:43:57 +0000 Subject: syncookies: update mss tables - ipv6 msstab: account for ipv6 header size - ipv4 msstab: add mss for Jumbograms. Signed-off-by: Florian Westphal Signed-off-by: David S. Miller --- net/ipv4/syncookies.c | 38 +++++++++++++++++++------------------- net/ipv6/syncookies.c | 39 +++++++++++++++------------------------ 2 files changed, 34 insertions(+), 43 deletions(-) (limited to 'net') diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c index c9dac86..a7cbcc4 100644 --- a/net/ipv4/syncookies.c +++ b/net/ipv4/syncookies.c @@ -138,23 +138,23 @@ static __u32 check_tcp_syn_cookie(__u32 cookie, __be32 saddr, __be32 daddr, } /* - * This table has to be sorted and terminated with (__u16)-1. - * XXX generate a better table. - * Unresolved Issues: HIPPI with a 64k MSS is not well supported. + * MSS Values are taken from the 2009 paper + * 'Measuring TCP Maximum Segment Size' by S. Alcock and R. Nelson: + * - values 1440 to 1460 accounted for 80% of observed mss values + * - values outside the 536-1460 range are rare (<0.2%). + * + * Table must be sorted. */ static __u16 const msstab[] = { - 64 - 1, - 256 - 1, - 512 - 1, - 536 - 1, - 1024 - 1, - 1440 - 1, - 1460 - 1, - 4312 - 1, - (__u16)-1 + 64, + 512, + 536, + 1024, + 1440, + 1460, + 4312, + 8960, }; -/* The number doesn't include the -1 terminator */ -#define NUM_MSS (ARRAY_SIZE(msstab) - 1) /* * Generate a syncookie. mssp points to the mss, which is returned @@ -169,10 +169,10 @@ __u32 cookie_v4_init_sequence(struct sock *sk, struct sk_buff *skb, __u16 *mssp) tcp_synq_overflow(sk); - /* XXX sort msstab[] by probability? Binary search? */ - for (mssind = 0; mss > msstab[mssind + 1]; mssind++) - ; - *mssp = msstab[mssind] + 1; + for (mssind = ARRAY_SIZE(msstab) - 1; mssind ; mssind--) + if (mss >= msstab[mssind]) + break; + *mssp = msstab[mssind]; NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESSENT); @@ -202,7 +202,7 @@ static inline int cookie_check(struct sk_buff *skb, __u32 cookie) jiffies / (HZ * 60), COUNTER_TRIES); - return mssind < NUM_MSS ? msstab[mssind] + 1 : 0; + return mssind < ARRAY_SIZE(msstab) ? msstab[mssind] : 0; } static inline struct sock *get_cookie_sock(struct sock *sk, struct sk_buff *skb, diff --git a/net/ipv6/syncookies.c b/net/ipv6/syncookies.c index 9fcb3ec..70d330f 100644 --- a/net/ipv6/syncookies.c +++ b/net/ipv6/syncookies.c @@ -27,28 +27,17 @@ extern __u32 syncookie_secret[2][16-4+SHA_DIGEST_WORDS]; #define COOKIEBITS 24 /* Upper bits store count */ #define COOKIEMASK (((__u32)1 << COOKIEBITS) - 1) -/* - * This table has to be sorted and terminated with (__u16)-1. - * XXX generate a better table. - * Unresolved Issues: HIPPI with a 64k MSS is not well supported. - * - * Taken directly from ipv4 implementation. - * Should this list be modified for ipv6 use or is it close enough? - * rfc 2460 8.3 suggests mss values 20 bytes less than ipv4 counterpart - */ +/* Table must be sorted. */ static __u16 const msstab[] = { - 64 - 1, - 256 - 1, - 512 - 1, - 536 - 1, - 1024 - 1, - 1440 - 1, - 1460 - 1, - 4312 - 1, - (__u16)-1 + 64, + 512, + 536, + 1280 - 60, + 1480 - 60, + 1500 - 60, + 4460 - 60, + 9000 - 60, }; -/* The number doesn't include the -1 terminator */ -#define NUM_MSS (ARRAY_SIZE(msstab) - 1) /* * This (misnamed) value is the age of syncookie which is permitted. @@ -134,9 +123,11 @@ __u32 cookie_v6_init_sequence(struct sock *sk, struct sk_buff *skb, __u16 *mssp) tcp_synq_overflow(sk); - for (mssind = 0; mss > msstab[mssind + 1]; mssind++) - ; - *mssp = msstab[mssind] + 1; + for (mssind = ARRAY_SIZE(msstab) - 1; mssind ; mssind--) + if (mss >= msstab[mssind]) + break; + + *mssp = msstab[mssind]; NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESSENT); @@ -154,7 +145,7 @@ static inline int cookie_check(struct sk_buff *skb, __u32 cookie) th->source, th->dest, seq, jiffies / (HZ * 60), COUNTER_TRIES); - return mssind < NUM_MSS ? msstab[mssind] + 1 : 0; + return mssind < ARRAY_SIZE(msstab) ? msstab[mssind] : 0; } struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb) -- cgit v1.1 From b78462ebc6a4ef9074aa80abebcdd470dc5f0ce0 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Wed, 2 Jun 2010 12:24:37 +0000 Subject: skbuff: add check for non-linear to warn_if_lro and needs_linearize We can avoid an unecessary cache miss by checking if the skb is non-linear before accessing gso_size/gso_type in skb_warn_if_lro, the same can also be done to avoid a cache miss on nr_frags if data_len is 0. Signed-off-by: Alexander Duyck Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- net/core/dev.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'net') diff --git a/net/core/dev.c b/net/core/dev.c index ec01a59..3abb3a6 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2103,9 +2103,10 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, static inline int skb_needs_linearize(struct sk_buff *skb, struct net_device *dev) { - return (skb_has_frags(skb) && !(dev->features & NETIF_F_FRAGLIST)) || - (skb_shinfo(skb)->nr_frags && (!(dev->features & NETIF_F_SG) || - illegal_highdma(dev, skb))); + return skb_is_nonlinear(skb) && + ((skb_has_frags(skb) && !(dev->features & NETIF_F_FRAGLIST)) || + (skb_shinfo(skb)->nr_frags && (!(dev->features & NETIF_F_SG) || + illegal_highdma(dev, skb)))); } /** -- cgit v1.1 From a8b690f98baf9fb1902b8eeab801351ea603fa3a Mon Sep 17 00:00:00 2001 From: Tom Herbert Date: Mon, 7 Jun 2010 00:43:42 -0700 Subject: tcp: Fix slowness in read /proc/net/tcp This patch address a serious performance issue in reading the TCP sockets table (/proc/net/tcp). Reading the full table is done by a number of sequential read operations. At each read operation, a seek is done to find the last socket that was previously read. This seek operation requires that the sockets in the table need to be counted up to the current file position, and to count each of these requires taking a lock for each non-empty bucket. The whole algorithm is O(n^2). The fix is to cache the last bucket value, offset within the bucket, and the file position returned by the last read operation. On the next sequential read, the bucket and offset are used to find the last read socket immediately without needing ot scan the previous buckets the table. This algorithm t read the whole table is O(n). The improvement offered by this patch is easily show by performing cat'ing /proc/net/tcp on a machine with a lot of connections. With about 182K connections in the table, I see the following: - Without patch time cat /proc/net/tcp > /dev/null real 1m56.729s user 0m0.214s sys 1m56.344s - With patch time cat /proc/net/tcp > /dev/null real 0m0.894s user 0m0.290s sys 0m0.594s Signed-off-by: Tom Herbert Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 92 ++++++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 84 insertions(+), 8 deletions(-) (limited to 'net') diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index acdc4c9..7f976af 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1980,6 +1980,11 @@ static inline struct inet_timewait_sock *tw_next(struct inet_timewait_sock *tw) hlist_nulls_entry(tw->tw_node.next, typeof(*tw), tw_node) : NULL; } +/* + * Get next listener socket follow cur. If cur is NULL, get first socket + * starting from bucket given in st->bucket; when st->bucket is zero the + * very first socket in the hash table is returned. + */ static void *listening_get_next(struct seq_file *seq, void *cur) { struct inet_connection_sock *icsk; @@ -1990,14 +1995,15 @@ static void *listening_get_next(struct seq_file *seq, void *cur) struct net *net = seq_file_net(seq); if (!sk) { - st->bucket = 0; - ilb = &tcp_hashinfo.listening_hash[0]; + ilb = &tcp_hashinfo.listening_hash[st->bucket]; spin_lock_bh(&ilb->lock); sk = sk_nulls_head(&ilb->head); + st->offset = 0; goto get_sk; } ilb = &tcp_hashinfo.listening_hash[st->bucket]; ++st->num; + ++st->offset; if (st->state == TCP_SEQ_STATE_OPENREQ) { struct request_sock *req = cur; @@ -2012,6 +2018,7 @@ static void *listening_get_next(struct seq_file *seq, void *cur) } req = req->dl_next; } + st->offset = 0; if (++st->sbucket >= icsk->icsk_accept_queue.listen_opt->nr_table_entries) break; get_req: @@ -2047,6 +2054,7 @@ start_req: read_unlock_bh(&icsk->icsk_accept_queue.syn_wait_lock); } spin_unlock_bh(&ilb->lock); + st->offset = 0; if (++st->bucket < INET_LHTABLE_SIZE) { ilb = &tcp_hashinfo.listening_hash[st->bucket]; spin_lock_bh(&ilb->lock); @@ -2060,7 +2068,12 @@ out: static void *listening_get_idx(struct seq_file *seq, loff_t *pos) { - void *rc = listening_get_next(seq, NULL); + struct tcp_iter_state *st = seq->private; + void *rc; + + st->bucket = 0; + st->offset = 0; + rc = listening_get_next(seq, NULL); while (rc && *pos) { rc = listening_get_next(seq, rc); @@ -2075,13 +2088,18 @@ static inline int empty_bucket(struct tcp_iter_state *st) hlist_nulls_empty(&tcp_hashinfo.ehash[st->bucket].twchain); } +/* + * Get first established socket starting from bucket given in st->bucket. + * If st->bucket is zero, the very first socket in the hash is returned. + */ static void *established_get_first(struct seq_file *seq) { struct tcp_iter_state *st = seq->private; struct net *net = seq_file_net(seq); void *rc = NULL; - for (st->bucket = 0; st->bucket <= tcp_hashinfo.ehash_mask; ++st->bucket) { + st->offset = 0; + for (; st->bucket <= tcp_hashinfo.ehash_mask; ++st->bucket) { struct sock *sk; struct hlist_nulls_node *node; struct inet_timewait_sock *tw; @@ -2126,6 +2144,7 @@ static void *established_get_next(struct seq_file *seq, void *cur) struct net *net = seq_file_net(seq); ++st->num; + ++st->offset; if (st->state == TCP_SEQ_STATE_TIME_WAIT) { tw = cur; @@ -2142,6 +2161,7 @@ get_tw: st->state = TCP_SEQ_STATE_ESTABLISHED; /* Look for next non empty bucket */ + st->offset = 0; while (++st->bucket <= tcp_hashinfo.ehash_mask && empty_bucket(st)) ; @@ -2169,7 +2189,11 @@ out: static void *established_get_idx(struct seq_file *seq, loff_t pos) { - void *rc = established_get_first(seq); + struct tcp_iter_state *st = seq->private; + void *rc; + + st->bucket = 0; + rc = established_get_first(seq); while (rc && pos) { rc = established_get_next(seq, rc); @@ -2194,24 +2218,72 @@ static void *tcp_get_idx(struct seq_file *seq, loff_t pos) return rc; } +static void *tcp_seek_last_pos(struct seq_file *seq) +{ + struct tcp_iter_state *st = seq->private; + int offset = st->offset; + int orig_num = st->num; + void *rc = NULL; + + switch (st->state) { + case TCP_SEQ_STATE_OPENREQ: + case TCP_SEQ_STATE_LISTENING: + if (st->bucket >= INET_LHTABLE_SIZE) + break; + st->state = TCP_SEQ_STATE_LISTENING; + rc = listening_get_next(seq, NULL); + while (offset-- && rc) + rc = listening_get_next(seq, rc); + if (rc) + break; + st->bucket = 0; + /* Fallthrough */ + case TCP_SEQ_STATE_ESTABLISHED: + case TCP_SEQ_STATE_TIME_WAIT: + st->state = TCP_SEQ_STATE_ESTABLISHED; + if (st->bucket > tcp_hashinfo.ehash_mask) + break; + rc = established_get_first(seq); + while (offset-- && rc) + rc = established_get_next(seq, rc); + } + + st->num = orig_num; + + return rc; +} + static void *tcp_seq_start(struct seq_file *seq, loff_t *pos) { struct tcp_iter_state *st = seq->private; + void *rc; + + if (*pos && *pos == st->last_pos) { + rc = tcp_seek_last_pos(seq); + if (rc) + goto out; + } + st->state = TCP_SEQ_STATE_LISTENING; st->num = 0; - return *pos ? tcp_get_idx(seq, *pos - 1) : SEQ_START_TOKEN; + st->bucket = 0; + st->offset = 0; + rc = *pos ? tcp_get_idx(seq, *pos - 1) : SEQ_START_TOKEN; + +out: + st->last_pos = *pos; + return rc; } static void *tcp_seq_next(struct seq_file *seq, void *v, loff_t *pos) { + struct tcp_iter_state *st = seq->private; void *rc = NULL; - struct tcp_iter_state *st; if (v == SEQ_START_TOKEN) { rc = tcp_get_idx(seq, 0); goto out; } - st = seq->private; switch (st->state) { case TCP_SEQ_STATE_OPENREQ: @@ -2219,6 +2291,8 @@ static void *tcp_seq_next(struct seq_file *seq, void *v, loff_t *pos) rc = listening_get_next(seq, v); if (!rc) { st->state = TCP_SEQ_STATE_ESTABLISHED; + st->bucket = 0; + st->offset = 0; rc = established_get_first(seq); } break; @@ -2229,6 +2303,7 @@ static void *tcp_seq_next(struct seq_file *seq, void *v, loff_t *pos) } out: ++*pos; + st->last_pos = *pos; return rc; } @@ -2267,6 +2342,7 @@ static int tcp_seq_open(struct inode *inode, struct file *file) s = ((struct seq_file *)file->private_data)->private; s->family = afinfo->family; + s->last_pos = 0; return 0; } -- cgit v1.1 From fe33147a58e7d1d3086bf823aabfd491d843be82 Mon Sep 17 00:00:00 2001 From: Alex Lorca Date: Mon, 7 Jun 2010 01:01:22 -0700 Subject: net-caif: Added missing lock validator constants CAIF is using "xxx-AF_MAX" strings for the lock validator. It should use its own strings. Signed-off-by: Alex Lorca Signed-off-by: David S. Miller --- net/core/sock.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'net') diff --git a/net/core/sock.c b/net/core/sock.c index 2cf7f9f..f9ce0db 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -156,7 +156,7 @@ static const char *const af_family_key_strings[AF_MAX+1] = { "sk_lock-27" , "sk_lock-28" , "sk_lock-AF_CAN" , "sk_lock-AF_TIPC" , "sk_lock-AF_BLUETOOTH", "sk_lock-IUCV" , "sk_lock-AF_RXRPC" , "sk_lock-AF_ISDN" , "sk_lock-AF_PHONET" , - "sk_lock-AF_IEEE802154", + "sk_lock-AF_IEEE802154", "sk_lock-AF_CAIF" , "sk_lock-AF_MAX" }; static const char *const af_family_slock_key_strings[AF_MAX+1] = { @@ -172,7 +172,7 @@ static const char *const af_family_slock_key_strings[AF_MAX+1] = { "slock-27" , "slock-28" , "slock-AF_CAN" , "slock-AF_TIPC" , "slock-AF_BLUETOOTH", "slock-AF_IUCV" , "slock-AF_RXRPC" , "slock-AF_ISDN" , "slock-AF_PHONET" , - "slock-AF_IEEE802154", + "slock-AF_IEEE802154", "slock-AF_CAIF" , "slock-AF_MAX" }; static const char *const af_family_clock_key_strings[AF_MAX+1] = { @@ -188,7 +188,7 @@ static const char *const af_family_clock_key_strings[AF_MAX+1] = { "clock-27" , "clock-28" , "clock-AF_CAN" , "clock-AF_TIPC" , "clock-AF_BLUETOOTH", "clock-AF_IUCV" , "clock-AF_RXRPC" , "clock-AF_ISDN" , "clock-AF_PHONET" , - "clock-AF_IEEE802154", + "clock-AF_IEEE802154", "clock-AF_CAIF" , "clock-AF_MAX" }; -- cgit v1.1 From 1789a640f55658d9a54c1868cc3405e4d85dbd8e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 3 Jun 2010 22:23:57 +0000 Subject: raw: avoid two atomics in xmit Avoid two atomic ops per raw_send_hdrinc() call Avoid two atomic ops per raw6_send_hdrinc() call Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/raw.c | 8 +++++--- net/ipv6/raw.c | 12 +++++++----- 2 files changed, 12 insertions(+), 8 deletions(-) (limited to 'net') diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 2c7a163..66cc3be 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -314,7 +314,7 @@ int raw_rcv(struct sock *sk, struct sk_buff *skb) } static int raw_send_hdrinc(struct sock *sk, void *from, size_t length, - struct rtable *rt, + struct rtable **rtp, unsigned int flags) { struct inet_sock *inet = inet_sk(sk); @@ -323,6 +323,7 @@ static int raw_send_hdrinc(struct sock *sk, void *from, size_t length, struct sk_buff *skb; unsigned int iphlen; int err; + struct rtable *rt = *rtp; if (length > rt->u.dst.dev->mtu) { ip_local_error(sk, EMSGSIZE, rt->rt_dst, inet->inet_dport, @@ -341,7 +342,8 @@ static int raw_send_hdrinc(struct sock *sk, void *from, size_t length, skb->priority = sk->sk_priority; skb->mark = sk->sk_mark; - skb_dst_set(skb, dst_clone(&rt->u.dst)); + skb_dst_set(skb, &rt->u.dst); + *rtp = NULL; skb_reset_network_header(skb); iph = ip_hdr(skb); @@ -576,7 +578,7 @@ back_from_confirm: if (inet->hdrincl) err = raw_send_hdrinc(sk, msg->msg_iov, len, - rt, msg->msg_flags); + &rt, msg->msg_flags); else { if (!ipc.addr) diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 864eb8e..968b964 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -602,13 +602,14 @@ out: } static int rawv6_send_hdrinc(struct sock *sk, void *from, int length, - struct flowi *fl, struct rt6_info *rt, + struct flowi *fl, struct dst_entry **dstp, unsigned int flags) { struct ipv6_pinfo *np = inet6_sk(sk); struct ipv6hdr *iph; struct sk_buff *skb; int err; + struct rt6_info *rt = (struct rt6_info *)*dstp; if (length > rt->u.dst.dev->mtu) { ipv6_local_error(sk, EMSGSIZE, fl, rt->u.dst.dev->mtu); @@ -626,7 +627,8 @@ static int rawv6_send_hdrinc(struct sock *sk, void *from, int length, skb->priority = sk->sk_priority; skb->mark = sk->sk_mark; - skb_dst_set(skb, dst_clone(&rt->u.dst)); + skb_dst_set(skb, &rt->u.dst); + *dstp = NULL; skb_put(skb, length); skb_reset_network_header(skb); @@ -886,9 +888,9 @@ static int rawv6_sendmsg(struct kiocb *iocb, struct sock *sk, goto do_confirm; back_from_confirm: - if (inet->hdrincl) { - err = rawv6_send_hdrinc(sk, msg->msg_iov, len, &fl, (struct rt6_info*)dst, msg->msg_flags); - } else { + if (inet->hdrincl) + err = rawv6_send_hdrinc(sk, msg->msg_iov, len, &fl, &dst, msg->msg_flags); + else { lock_sock(sk); err = ip6_append_data(sk, ip_generic_getfrag, msg->msg_iov, len, 0, hlimit, tclass, opt, &fl, (struct rt6_info*)dst, -- cgit v1.1 From f2a03367c072150c881fa23ce3d3f76b8236018f Mon Sep 17 00:00:00 2001 From: Changli Gao Date: Fri, 4 Jun 2010 01:33:33 +0000 Subject: htb: remove two unnecessary assignments remove two unnecessary assignments we don't need to assign NULL when initialize structure objects. Signed-off-by: Changli Gao ---- net/sched/sch_htb.c | 2 -- 1 file changed, 2 deletions(-) Signed-off-by: David S. Miller --- net/sched/sch_htb.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'net') diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c index 0b52b8de..4be8d04 100644 --- a/net/sched/sch_htb.c +++ b/net/sched/sch_htb.c @@ -1550,7 +1550,6 @@ static const struct Qdisc_class_ops htb_class_ops = { }; static struct Qdisc_ops htb_qdisc_ops __read_mostly = { - .next = NULL, .cl_ops = &htb_class_ops, .id = "htb", .priv_size = sizeof(struct htb_sched), @@ -1561,7 +1560,6 @@ static struct Qdisc_ops htb_qdisc_ops __read_mostly = { .init = htb_init, .reset = htb_reset, .destroy = htb_destroy, - .change = NULL /* htb_change */, .dump = htb_dump, .owner = THIS_MODULE, }; -- cgit v1.1 From 9dacaf17a60101a55d456cc7b00e269d8145aa0d Mon Sep 17 00:00:00 2001 From: jamal Date: Fri, 4 Jun 2010 02:43:06 +0000 Subject: net sched: make pedit check for clones instead Now that the core path doesnt set OK to munge we detect writable skbs by looking to see if they are cloned. Signed-off-by: Jamal Hadi Salim Signed-off-by: David S. Miller --- net/sched/act_pedit.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'net') diff --git a/net/sched/act_pedit.c b/net/sched/act_pedit.c index 50e3d94..a0593c9 100644 --- a/net/sched/act_pedit.c +++ b/net/sched/act_pedit.c @@ -127,8 +127,7 @@ static int tcf_pedit(struct sk_buff *skb, struct tc_action *a, int i, munged = 0; unsigned int off; - if (!(skb->tc_verd & TC_OK2MUNGE)) { - /* should we set skb->cloned? */ + if (skb_cloned(skb)) { if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC)) { return p->tcf_action; } -- cgit v1.1 From 271c1dfa61bc90a57648ff96f3eb92d4b4d4f11e Mon Sep 17 00:00:00 2001 From: jamal Date: Fri, 4 Jun 2010 02:06:22 +0000 Subject: net: Remove unnecessary net action assertion The extra assertion to allow packet munging only when there are no other ptypes listening which may have worked around an old bug is unnecessary. It is sufficient to check if the skb is cloned before trampling on it. Thanks to Herbert Xu for being persistent and patient in getting this across. [Note that cloning checks and assertions are the general rule used by tc actions (documentation/networking/tc-actions-env-rules.txt)]. Signed-off-by: Jamal Hadi Salim Signed-off-by: David S. Miller --- net/core/dev.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'net') diff --git a/net/core/dev.c b/net/core/dev.c index b65347c..c8d1277 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2663,9 +2663,6 @@ static inline struct sk_buff *handle_ing(struct sk_buff *skb, if (*pt_prev) { *ret = deliver_skb(skb, *pt_prev, orig_dev); *pt_prev = NULL; - } else { - /* Huh? Why does turning on AF_PACKET affect this? */ - skb->tc_verd = SET_TC_OK2MUNGE(skb->tc_verd); } switch (ing_filter(skb)) { -- cgit v1.1 From 11b7c60988e5fbabb4e150612931cc068559af16 Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Mon, 7 Jun 2010 15:02:17 -0400 Subject: mac80211: fix lock leak w/ ARP filtering and w/o CONFIG_INET "mac80211: make ARP filtering depend on CONFIG_INET" introduced this potential locking leak. Reported-by: Juuso Oikarinen Signed-off-by: John W. Linville --- net/mac80211/mlme.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'net') diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 3623bb7..9420cf1 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2078,10 +2078,9 @@ static enum work_done_result ieee80211_assoc_done(struct ieee80211_work *wk, cfg80211_send_assoc_timeout(wk->sdata->dev, wk->filter_ta); return WORK_DONE_DESTROY; -#ifdef CONFIG_INET } else { mutex_unlock(&wk->sdata->u.mgd.mtx); - +#ifdef CONFIG_INET /* * configure ARP filter IP addresses to the driver, * intentionally outside the mgd mutex. -- cgit v1.1 From 66018506e15bea62de4eefc3298f170b4bfcf5ef Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 7 Jun 2010 03:12:08 +0000 Subject: ip: Router Alert RCU conversion Straightforward conversion to RCU. One rwlock becomes a spinlock, and is static. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/ip_input.c | 11 +++-------- net/ipv4/ip_sockglue.c | 23 ++++++++++++++--------- 2 files changed, 17 insertions(+), 17 deletions(-) (limited to 'net') diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c index d52c9da..d274078 100644 --- a/net/ipv4/ip_input.c +++ b/net/ipv4/ip_input.c @@ -146,7 +146,7 @@ #include /* - * Process Router Attention IP option + * Process Router Attention IP option (RFC 2113) */ int ip_call_ra_chain(struct sk_buff *skb) { @@ -155,8 +155,7 @@ int ip_call_ra_chain(struct sk_buff *skb) struct sock *last = NULL; struct net_device *dev = skb->dev; - read_lock(&ip_ra_lock); - for (ra = ip_ra_chain; ra; ra = ra->next) { + for (ra = rcu_dereference(ip_ra_chain); ra; ra = rcu_dereference(ra->next)) { struct sock *sk = ra->sk; /* If socket is bound to an interface, only report @@ -167,10 +166,8 @@ int ip_call_ra_chain(struct sk_buff *skb) sk->sk_bound_dev_if == dev->ifindex) && net_eq(sock_net(sk), dev_net(dev))) { if (ip_hdr(skb)->frag_off & htons(IP_MF | IP_OFFSET)) { - if (ip_defrag(skb, IP_DEFRAG_CALL_RA_CHAIN)) { - read_unlock(&ip_ra_lock); + if (ip_defrag(skb, IP_DEFRAG_CALL_RA_CHAIN)) return 1; - } } if (last) { struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC); @@ -183,10 +180,8 @@ int ip_call_ra_chain(struct sk_buff *skb) if (last) { raw_rcv(last, skb); - read_unlock(&ip_ra_lock); return 1; } - read_unlock(&ip_ra_lock); return 0; } diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c index ce23178..08b9519 100644 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@ -239,7 +239,12 @@ int ip_cmsg_send(struct net *net, struct msghdr *msg, struct ipcm_cookie *ipc) sent to multicast group to reach destination designated router. */ struct ip_ra_chain *ip_ra_chain; -DEFINE_RWLOCK(ip_ra_lock); +static DEFINE_SPINLOCK(ip_ra_lock); + +static void ip_ra_free_rcu(struct rcu_head *head) +{ + kfree(container_of(head, struct ip_ra_chain, rcu)); +} int ip_ra_control(struct sock *sk, unsigned char on, void (*destructor)(struct sock *)) @@ -251,35 +256,35 @@ int ip_ra_control(struct sock *sk, unsigned char on, new_ra = on ? kmalloc(sizeof(*new_ra), GFP_KERNEL) : NULL; - write_lock_bh(&ip_ra_lock); + spin_lock_bh(&ip_ra_lock); for (rap = &ip_ra_chain; (ra = *rap) != NULL; rap = &ra->next) { if (ra->sk == sk) { if (on) { - write_unlock_bh(&ip_ra_lock); + spin_unlock_bh(&ip_ra_lock); kfree(new_ra); return -EADDRINUSE; } - *rap = ra->next; - write_unlock_bh(&ip_ra_lock); + rcu_assign_pointer(*rap, ra->next); + spin_unlock_bh(&ip_ra_lock); if (ra->destructor) ra->destructor(sk); sock_put(sk); - kfree(ra); + call_rcu(&ra->rcu, ip_ra_free_rcu); return 0; } } if (new_ra == NULL) { - write_unlock_bh(&ip_ra_lock); + spin_unlock_bh(&ip_ra_lock); return -ENOBUFS; } new_ra->sk = sk; new_ra->destructor = destructor; new_ra->next = ra; - *rap = new_ra; + rcu_assign_pointer(*rap, new_ra); sock_hold(sk); - write_unlock_bh(&ip_ra_lock); + spin_unlock_bh(&ip_ra_lock); return 0; } -- cgit v1.1 From 9a57a9d291980302b4a3184fbc47dbddac71903e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 7 Jun 2010 03:17:10 +0000 Subject: igmp: avoid two atomic ops in igmp_rcv() in_dev_get() -> __in_dev_get_rcu() in a rcu protected function. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/igmp.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) (limited to 'net') diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index 250cb5e..3294f54 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -916,18 +916,19 @@ static void igmp_heard_query(struct in_device *in_dev, struct sk_buff *skb, read_unlock(&in_dev->mc_list_lock); } +/* called in rcu_read_lock() section */ int igmp_rcv(struct sk_buff *skb) { /* This basically follows the spec line by line -- see RFC1112 */ struct igmphdr *ih; - struct in_device *in_dev = in_dev_get(skb->dev); + struct in_device *in_dev = __in_dev_get_rcu(skb->dev); int len = skb->len; if (in_dev == NULL) goto drop; if (!pskb_may_pull(skb, sizeof(struct igmphdr))) - goto drop_ref; + goto drop; switch (skb->ip_summed) { case CHECKSUM_COMPLETE: @@ -937,7 +938,7 @@ int igmp_rcv(struct sk_buff *skb) case CHECKSUM_NONE: skb->csum = 0; if (__skb_checksum_complete(skb)) - goto drop_ref; + goto drop; } ih = igmp_hdr(skb); @@ -957,7 +958,6 @@ int igmp_rcv(struct sk_buff *skb) break; case IGMP_PIM: #ifdef CONFIG_IP_PIMSM_V1 - in_dev_put(in_dev); return pim_rcv_v1(skb); #endif case IGMPV3_HOST_MEMBERSHIP_REPORT: @@ -971,8 +971,6 @@ int igmp_rcv(struct sk_buff *skb) break; } -drop_ref: - in_dev_put(in_dev); drop: kfree_skb(skb); return 0; -- cgit v1.1 From ed7865a47d4759e85bbd7bed44bee411d868eaad Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 7 Jun 2010 21:49:44 -0700 Subject: ipv4: avoid two atomic ops in ip_rt_redirect() in_dev_get() -> __in_dev_get_rcu() in a rcu protected function. [ Fix build with CONFIG_IP_ROUTE_VERBOSE disabled. -DaveM ] Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/route.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) (limited to 'net') diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 7b8eacd..883b5c7 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -1343,11 +1343,12 @@ static void rt_del(unsigned hash, struct rtable *rt) spin_unlock_bh(rt_hash_lock_addr(hash)); } +/* called in rcu_read_lock() section */ void ip_rt_redirect(__be32 old_gw, __be32 daddr, __be32 new_gw, __be32 saddr, struct net_device *dev) { int i, k; - struct in_device *in_dev = in_dev_get(dev); + struct in_device *in_dev = __in_dev_get_rcu(dev); struct rtable *rth, **rthp; __be32 skeys[2] = { saddr, 0 }; int ikeys[2] = { dev->ifindex, 0 }; @@ -1383,7 +1384,6 @@ void ip_rt_redirect(__be32 old_gw, __be32 daddr, __be32 new_gw, rthp=&rt_hash_table[hash].chain; - rcu_read_lock(); while ((rth = rcu_dereference(*rthp)) != NULL) { struct rtable *rt; @@ -1405,12 +1405,10 @@ void ip_rt_redirect(__be32 old_gw, __be32 daddr, __be32 new_gw, break; dst_hold(&rth->u.dst); - rcu_read_unlock(); rt = dst_alloc(&ipv4_dst_ops); if (rt == NULL) { ip_rt_put(rth); - in_dev_put(in_dev); return; } @@ -1463,12 +1461,10 @@ void ip_rt_redirect(__be32 old_gw, __be32 daddr, __be32 new_gw, ip_rt_put(rt); goto do_next; } - rcu_read_unlock(); do_next: ; } } - in_dev_put(in_dev); return; reject_redirect: @@ -1479,7 +1475,7 @@ reject_redirect: &old_gw, dev->name, &new_gw, &saddr, &daddr); #endif - in_dev_put(in_dev); + ; } static struct dst_entry *ipv4_negative_advice(struct dst_entry *dst) -- cgit v1.1 From 6e8b11b43b0c2e901734e2cdd70c6e325a90c4ef Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 7 Jun 2010 03:54:46 +0000 Subject: net: avoid two atomic ops in ip_rcv_options() in_dev_get() -> __in_dev_get_rcu() in a rcu protected function. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/ip_input.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c index d274078..08a3b12 100644 --- a/net/ipv4/ip_input.c +++ b/net/ipv4/ip_input.c @@ -293,18 +293,16 @@ static inline int ip_rcv_options(struct sk_buff *skb) } if (unlikely(opt->srr)) { - struct in_device *in_dev = in_dev_get(dev); + struct in_device *in_dev = __in_dev_get_rcu(dev); + if (in_dev) { if (!IN_DEV_SOURCE_ROUTE(in_dev)) { if (IN_DEV_LOG_MARTIANS(in_dev) && net_ratelimit()) printk(KERN_INFO "source route option %pI4 -> %pI4\n", &iph->saddr, &iph->daddr); - in_dev_put(in_dev); goto drop; } - - in_dev_put(in_dev); } if (ip_options_rcv_srr(skb)) -- cgit v1.1 From bb69ae049fcc986fcd742eb90ca0d44a7a49c9f1 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 7 Jun 2010 11:42:13 +0000 Subject: anycast: Some RCU conversions - dev_get_by_flags() changed to dev_get_by_flags_rcu() - ipv6_sock_ac_join() dont touch dev & idev refcounts - ipv6_sock_ac_drop() dont touch dev & idev refcounts - ipv6_sock_ac_close() dont touch dev & idev refcounts - ipv6_dev_ac_dec() dount touch idev refcount - ipv6_chk_acast_addr() dont touch idev refcount Signed-off-by: Eric Dumazet CC: Hideaki YOSHIFUJI Signed-off-by: David S. Miller --- net/core/dev.c | 14 +++------ net/ipv6/anycast.c | 90 +++++++++++++++++++++++++----------------------------- 2 files changed, 47 insertions(+), 57 deletions(-) (limited to 'net') diff --git a/net/core/dev.c b/net/core/dev.c index c8d1277..6f330ce 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -803,35 +803,31 @@ struct net_device *dev_getfirstbyhwtype(struct net *net, unsigned short type) EXPORT_SYMBOL(dev_getfirstbyhwtype); /** - * dev_get_by_flags - find any device with given flags + * dev_get_by_flags_rcu - find any device with given flags * @net: the applicable net namespace * @if_flags: IFF_* values * @mask: bitmask of bits in if_flags to check * * Search for any interface with the given flags. Returns NULL if a device - * is not found or a pointer to the device. The device returned has - * had a reference added and the pointer is safe until the user calls - * dev_put to indicate they have finished with it. + * is not found or a pointer to the device. Must be called inside + * rcu_read_lock(), and result refcount is unchanged. */ -struct net_device *dev_get_by_flags(struct net *net, unsigned short if_flags, +struct net_device *dev_get_by_flags_rcu(struct net *net, unsigned short if_flags, unsigned short mask) { struct net_device *dev, *ret; ret = NULL; - rcu_read_lock(); for_each_netdev_rcu(net, dev) { if (((dev->flags ^ if_flags) & mask) == 0) { - dev_hold(dev); ret = dev; break; } } - rcu_read_unlock(); return ret; } -EXPORT_SYMBOL(dev_get_by_flags); +EXPORT_SYMBOL(dev_get_by_flags_rcu); /** * dev_valid_name - check if name is okay for network device diff --git a/net/ipv6/anycast.c b/net/ipv6/anycast.c index b5b0705..f058fbd 100644 --- a/net/ipv6/anycast.c +++ b/net/ipv6/anycast.c @@ -77,41 +77,40 @@ int ipv6_sock_ac_join(struct sock *sk, int ifindex, struct in6_addr *addr) pac->acl_next = NULL; ipv6_addr_copy(&pac->acl_addr, addr); + rcu_read_lock(); if (ifindex == 0) { struct rt6_info *rt; rt = rt6_lookup(net, addr, NULL, 0, 0); if (rt) { dev = rt->rt6i_dev; - dev_hold(dev); dst_release(&rt->u.dst); } else if (ishost) { err = -EADDRNOTAVAIL; - goto out_free_pac; + goto error; } else { /* router, no matching interface: just pick one */ - - dev = dev_get_by_flags(net, IFF_UP, IFF_UP|IFF_LOOPBACK); + dev = dev_get_by_flags_rcu(net, IFF_UP, + IFF_UP | IFF_LOOPBACK); } } else - dev = dev_get_by_index(net, ifindex); + dev = dev_get_by_index_rcu(net, ifindex); if (dev == NULL) { err = -ENODEV; - goto out_free_pac; + goto error; } - idev = in6_dev_get(dev); + idev = __in6_dev_get(dev); if (!idev) { if (ifindex) err = -ENODEV; else err = -EADDRNOTAVAIL; - goto out_dev_put; + goto error; } /* reset ishost, now that we have a specific device */ ishost = !idev->cnf.forwarding; - in6_dev_put(idev); pac->acl_ifindex = dev->ifindex; @@ -124,26 +123,22 @@ int ipv6_sock_ac_join(struct sock *sk, int ifindex, struct in6_addr *addr) if (ishost) err = -EADDRNOTAVAIL; if (err) - goto out_dev_put; + goto error; } err = ipv6_dev_ac_inc(dev, addr); - if (err) - goto out_dev_put; - - write_lock_bh(&ipv6_sk_ac_lock); - pac->acl_next = np->ipv6_ac_list; - np->ipv6_ac_list = pac; - write_unlock_bh(&ipv6_sk_ac_lock); - - dev_put(dev); - - return 0; + if (!err) { + write_lock_bh(&ipv6_sk_ac_lock); + pac->acl_next = np->ipv6_ac_list; + np->ipv6_ac_list = pac; + write_unlock_bh(&ipv6_sk_ac_lock); + pac = NULL; + } -out_dev_put: - dev_put(dev); -out_free_pac: - sock_kfree_s(sk, pac, sizeof(*pac)); +error: + rcu_read_unlock(); + if (pac) + sock_kfree_s(sk, pac, sizeof(*pac)); return err; } @@ -176,11 +171,12 @@ int ipv6_sock_ac_drop(struct sock *sk, int ifindex, struct in6_addr *addr) write_unlock_bh(&ipv6_sk_ac_lock); - dev = dev_get_by_index(net, pac->acl_ifindex); - if (dev) { + rcu_read_lock(); + dev = dev_get_by_index_rcu(net, pac->acl_ifindex); + if (dev) ipv6_dev_ac_dec(dev, &pac->acl_addr); - dev_put(dev); - } + rcu_read_unlock(); + sock_kfree_s(sk, pac, sizeof(*pac)); return 0; } @@ -199,13 +195,12 @@ void ipv6_sock_ac_close(struct sock *sk) write_unlock_bh(&ipv6_sk_ac_lock); prev_index = 0; + rcu_read_lock(); while (pac) { struct ipv6_ac_socklist *next = pac->acl_next; if (pac->acl_ifindex != prev_index) { - if (dev) - dev_put(dev); - dev = dev_get_by_index(net, pac->acl_ifindex); + dev = dev_get_by_index_rcu(net, pac->acl_ifindex); prev_index = pac->acl_ifindex; } if (dev) @@ -213,8 +208,7 @@ void ipv6_sock_ac_close(struct sock *sk) sock_kfree_s(sk, pac, sizeof(*pac)); pac = next; } - if (dev) - dev_put(dev); + rcu_read_unlock(); } #if 0 @@ -363,33 +357,32 @@ int __ipv6_dev_ac_dec(struct inet6_dev *idev, struct in6_addr *addr) return 0; } +/* called with rcu_read_lock() */ static int ipv6_dev_ac_dec(struct net_device *dev, struct in6_addr *addr) { - int ret; - struct inet6_dev *idev = in6_dev_get(dev); + struct inet6_dev *idev = __in6_dev_get(dev); + if (idev == NULL) return -ENODEV; - ret = __ipv6_dev_ac_dec(idev, addr); - in6_dev_put(idev); - return ret; + return __ipv6_dev_ac_dec(idev, addr); } /* * check if the interface has this anycast address + * called with rcu_read_lock() */ static int ipv6_chk_acast_dev(struct net_device *dev, struct in6_addr *addr) { struct inet6_dev *idev; struct ifacaddr6 *aca; - idev = in6_dev_get(dev); + idev = __in6_dev_get(dev); if (idev) { read_lock_bh(&idev->lock); for (aca = idev->ac_list; aca; aca = aca->aca_next) if (ipv6_addr_equal(&aca->aca_addr, addr)) break; read_unlock_bh(&idev->lock); - in6_dev_put(idev); return aca != NULL; } return 0; @@ -403,14 +396,15 @@ int ipv6_chk_acast_addr(struct net *net, struct net_device *dev, { int found = 0; - if (dev) - return ipv6_chk_acast_dev(dev, addr); rcu_read_lock(); - for_each_netdev_rcu(net, dev) - if (ipv6_chk_acast_dev(dev, addr)) { - found = 1; - break; - } + if (dev) + found = ipv6_chk_acast_dev(dev, addr); + else + for_each_netdev_rcu(net, dev) + if (ipv6_chk_acast_dev(dev, addr)) { + found = 1; + break; + } rcu_read_unlock(); return found; } -- cgit v1.1 From 90b726097ba0dcc1f9725182513e669a30e77db5 Mon Sep 17 00:00:00 2001 From: Juuso Oikarinen Date: Mon, 7 Jun 2010 10:52:12 +0300 Subject: mac80211: Add netif state checking to ieee80211_ifa_changed There's a window for ieee80211_ifa_changed() to get called whilst the managed mode mutex has not been initialized when opening and stopping the interface. Currently this causes a kernel BUG like the following: [ 132.460013] kernel BUG at /home/wifi/iwlwifi-2.6/net/mac80211/main.c:380! [ 132.460013] invalid opcode: 0000 [#1] SMP The mutex is initialized during open(), hence once netif_running() is true, the mutex should be valid. Fix by adding a netif_running() check to the function. Reported-by: Reinette Chatre Signed-off-by: Juuso Oikarinen Tested-by: Reinette Chatre Signed-off-by: John W. Linville --- net/mac80211/main.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'net') diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 5706156..88b671a 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -359,6 +359,9 @@ static int ieee80211_ifa_changed(struct notifier_block *nb, struct ieee80211_sub_if_data *sdata; struct ieee80211_if_managed *ifmgd; + if (!netif_running(ndev)) + return NOTIFY_DONE; + /* Make sure it's our interface that got changed */ if (!wdev) return NOTIFY_DONE; -- cgit v1.1 From 5bfddbd46a95c978f4d3c992339cbdf4f4b790a3 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 8 Jun 2010 16:09:52 +0200 Subject: netfilter: nf_conntrack: IPS_UNTRACKED bit NOTRACK makes all cpus share a cache line on nf_conntrack_untracked twice per packet. This is bad for performance. __read_mostly annotation is also a bad choice. This patch introduces IPS_UNTRACKED bit so that we can use later a per_cpu untrack structure more easily. A new helper, nf_ct_untracked_get() returns a pointer to nf_conntrack_untracked. Another one, nf_ct_untracked_status_or() is used by nf_nat_init() to add IPS_NAT_DONE_MASK bits to untracked status. nf_ct_is_untracked() prototype is changed to work on a nf_conn pointer. Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/ipv4/netfilter/nf_nat_core.c | 2 +- net/ipv4/netfilter/nf_nat_standalone.c | 2 +- net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c | 2 +- net/netfilter/nf_conntrack_core.c | 11 ++++++++--- net/netfilter/nf_conntrack_netlink.c | 2 +- net/netfilter/xt_CT.c | 4 ++-- net/netfilter/xt_NOTRACK.c | 2 +- net/netfilter/xt_TEE.c | 4 ++-- net/netfilter/xt_cluster.c | 2 +- net/netfilter/xt_conntrack.c | 11 ++++++----- net/netfilter/xt_socket.c | 2 +- net/netfilter/xt_state.c | 14 ++++++++------ 12 files changed, 33 insertions(+), 25 deletions(-) (limited to 'net') diff --git a/net/ipv4/netfilter/nf_nat_core.c b/net/ipv4/netfilter/nf_nat_core.c index 4f8bddb..c7719b2 100644 --- a/net/ipv4/netfilter/nf_nat_core.c +++ b/net/ipv4/netfilter/nf_nat_core.c @@ -742,7 +742,7 @@ static int __init nf_nat_init(void) spin_unlock_bh(&nf_nat_lock); /* Initialize fake conntrack so that NAT will skip it */ - nf_conntrack_untracked.status |= IPS_NAT_DONE_MASK; + nf_ct_untracked_status_or(IPS_NAT_DONE_MASK); l3proto = nf_ct_l3proto_find_get((u_int16_t)AF_INET); diff --git a/net/ipv4/netfilter/nf_nat_standalone.c b/net/ipv4/netfilter/nf_nat_standalone.c index beb2581..6723c68 100644 --- a/net/ipv4/netfilter/nf_nat_standalone.c +++ b/net/ipv4/netfilter/nf_nat_standalone.c @@ -98,7 +98,7 @@ nf_nat_fn(unsigned int hooknum, return NF_ACCEPT; /* Don't try to NAT if this packet is not conntracked */ - if (ct == &nf_conntrack_untracked) + if (nf_ct_is_untracked(ct)) return NF_ACCEPT; nat = nfct_nat(ct); diff --git a/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c b/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c index 9be8177..1df3c8b 100644 --- a/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c +++ b/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c @@ -208,7 +208,7 @@ icmpv6_error(struct net *net, struct nf_conn *tmpl, type = icmp6h->icmp6_type - 130; if (type >= 0 && type < sizeof(noct_valid_new) && noct_valid_new[type]) { - skb->nfct = &nf_conntrack_untracked.ct_general; + skb->nfct = &nf_ct_untracked_get()->ct_general; skb->nfctinfo = IP_CT_NEW; nf_conntrack_get(skb->nfct); return NF_ACCEPT; diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c index eeeb8bc..6c1da21 100644 --- a/net/netfilter/nf_conntrack_core.c +++ b/net/netfilter/nf_conntrack_core.c @@ -62,7 +62,7 @@ EXPORT_SYMBOL_GPL(nf_conntrack_htable_size); unsigned int nf_conntrack_max __read_mostly; EXPORT_SYMBOL_GPL(nf_conntrack_max); -struct nf_conn nf_conntrack_untracked __read_mostly; +struct nf_conn nf_conntrack_untracked; EXPORT_SYMBOL_GPL(nf_conntrack_untracked); static int nf_conntrack_hash_rnd_initted; @@ -1321,6 +1321,12 @@ EXPORT_SYMBOL_GPL(nf_conntrack_set_hashsize); module_param_call(hashsize, nf_conntrack_set_hashsize, param_get_uint, &nf_conntrack_htable_size, 0600); +void nf_ct_untracked_status_or(unsigned long bits) +{ + nf_conntrack_untracked.status |= bits; +} +EXPORT_SYMBOL_GPL(nf_ct_untracked_status_or); + static int nf_conntrack_init_init_net(void) { int max_factor = 8; @@ -1368,8 +1374,7 @@ static int nf_conntrack_init_init_net(void) #endif atomic_set(&nf_conntrack_untracked.ct_general.use, 1); /* - and look it like as a confirmed connection */ - set_bit(IPS_CONFIRMED_BIT, &nf_conntrack_untracked.status); - + nf_ct_untracked_status_or(IPS_CONFIRMED | IPS_UNTRACKED); return 0; #ifdef CONFIG_NF_CONNTRACK_ZONES diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index c42ff6a..5bae1cd 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -480,7 +480,7 @@ ctnetlink_conntrack_event(unsigned int events, struct nf_ct_event *item) int err; /* ignore our fake conntrack entry */ - if (ct == &nf_conntrack_untracked) + if (nf_ct_is_untracked(ct)) return 0; if (events & (1 << IPCT_DESTROY)) { diff --git a/net/netfilter/xt_CT.c b/net/netfilter/xt_CT.c index 562bf32..0cb6053 100644 --- a/net/netfilter/xt_CT.c +++ b/net/netfilter/xt_CT.c @@ -67,7 +67,7 @@ static int xt_ct_tg_check(const struct xt_tgchk_param *par) return -EINVAL; if (info->flags & XT_CT_NOTRACK) { - ct = &nf_conntrack_untracked; + ct = nf_ct_untracked_get(); atomic_inc(&ct->ct_general.use); goto out; } @@ -132,7 +132,7 @@ static void xt_ct_tg_destroy(const struct xt_tgdtor_param *par) struct nf_conn *ct = info->ct; struct nf_conn_help *help; - if (ct != &nf_conntrack_untracked) { + if (!nf_ct_is_untracked(ct)) { help = nfct_help(ct); if (help) module_put(help->helper->me); diff --git a/net/netfilter/xt_NOTRACK.c b/net/netfilter/xt_NOTRACK.c index 512b912..9d78218 100644 --- a/net/netfilter/xt_NOTRACK.c +++ b/net/netfilter/xt_NOTRACK.c @@ -23,7 +23,7 @@ notrack_tg(struct sk_buff *skb, const struct xt_action_param *par) If there is a real ct entry correspondig to this packet, it'll hang aroun till timing out. We don't deal with it for performance reasons. JK */ - skb->nfct = &nf_conntrack_untracked.ct_general; + skb->nfct = &nf_ct_untracked_get()->ct_general; skb->nfctinfo = IP_CT_NEW; nf_conntrack_get(skb->nfct); diff --git a/net/netfilter/xt_TEE.c b/net/netfilter/xt_TEE.c index 859d9fd..7a11826 100644 --- a/net/netfilter/xt_TEE.c +++ b/net/netfilter/xt_TEE.c @@ -104,7 +104,7 @@ tee_tg4(struct sk_buff *skb, const struct xt_action_param *par) #ifdef WITH_CONNTRACK /* Avoid counting cloned packets towards the original connection. */ nf_conntrack_put(skb->nfct); - skb->nfct = &nf_conntrack_untracked.ct_general; + skb->nfct = &nf_ct_untracked_get()->ct_general; skb->nfctinfo = IP_CT_NEW; nf_conntrack_get(skb->nfct); #endif @@ -177,7 +177,7 @@ tee_tg6(struct sk_buff *skb, const struct xt_action_param *par) #ifdef WITH_CONNTRACK nf_conntrack_put(skb->nfct); - skb->nfct = &nf_conntrack_untracked.ct_general; + skb->nfct = &nf_ct_untracked_get()->ct_general; skb->nfctinfo = IP_CT_NEW; nf_conntrack_get(skb->nfct); #endif diff --git a/net/netfilter/xt_cluster.c b/net/netfilter/xt_cluster.c index 30b95a1..f4af1bf 100644 --- a/net/netfilter/xt_cluster.c +++ b/net/netfilter/xt_cluster.c @@ -120,7 +120,7 @@ xt_cluster_mt(const struct sk_buff *skb, struct xt_action_param *par) if (ct == NULL) return false; - if (ct == &nf_conntrack_untracked) + if (nf_ct_is_untracked(ct)) return false; if (ct->master) diff --git a/net/netfilter/xt_conntrack.c b/net/netfilter/xt_conntrack.c index 39681f1..e536710 100644 --- a/net/netfilter/xt_conntrack.c +++ b/net/netfilter/xt_conntrack.c @@ -123,11 +123,12 @@ conntrack_mt(const struct sk_buff *skb, struct xt_action_param *par, ct = nf_ct_get(skb, &ctinfo); - if (ct == &nf_conntrack_untracked) - statebit = XT_CONNTRACK_STATE_UNTRACKED; - else if (ct != NULL) - statebit = XT_CONNTRACK_STATE_BIT(ctinfo); - else + if (ct) { + if (nf_ct_is_untracked(ct)) + statebit = XT_CONNTRACK_STATE_UNTRACKED; + else + statebit = XT_CONNTRACK_STATE_BIT(ctinfo); + } else statebit = XT_CONNTRACK_STATE_INVALID; if (info->match_flags & XT_CONNTRACK_STATE) { diff --git a/net/netfilter/xt_socket.c b/net/netfilter/xt_socket.c index 3d54c23..1ca8990 100644 --- a/net/netfilter/xt_socket.c +++ b/net/netfilter/xt_socket.c @@ -127,7 +127,7 @@ socket_match(const struct sk_buff *skb, struct xt_action_param *par, * reply packet of an established SNAT-ted connection. */ ct = nf_ct_get(skb, &ctinfo); - if (ct && (ct != &nf_conntrack_untracked) && + if (ct && !nf_ct_is_untracked(ct) && ((iph->protocol != IPPROTO_ICMP && ctinfo == IP_CT_IS_REPLY + IP_CT_ESTABLISHED) || (iph->protocol == IPPROTO_ICMP && diff --git a/net/netfilter/xt_state.c b/net/netfilter/xt_state.c index e12e053..a507922 100644 --- a/net/netfilter/xt_state.c +++ b/net/netfilter/xt_state.c @@ -26,14 +26,16 @@ state_mt(const struct sk_buff *skb, struct xt_action_param *par) const struct xt_state_info *sinfo = par->matchinfo; enum ip_conntrack_info ctinfo; unsigned int statebit; + struct nf_conn *ct = nf_ct_get(skb, &ctinfo); - if (nf_ct_is_untracked(skb)) - statebit = XT_STATE_UNTRACKED; - else if (!nf_ct_get(skb, &ctinfo)) + if (!ct) statebit = XT_STATE_INVALID; - else - statebit = XT_STATE_BIT(ctinfo); - + else { + if (nf_ct_is_untracked(ct)) + statebit = XT_STATE_UNTRACKED; + else + statebit = XT_STATE_BIT(ctinfo); + } return (sinfo->statemask & statebit); } -- cgit v1.1 From b3c5163fe0193a74016dba1bb22491e0d1e9aaa4 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 9 Jun 2010 14:43:38 +0200 Subject: netfilter: nf_conntrack: per_cpu untracking NOTRACK makes all cpus share a cache line on nf_conntrack_untracked twice per packet, slowing down performance. This patch converts it to a per_cpu variable. Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/netfilter/nf_conntrack_core.c | 36 ++++++++++++++++++++++++++---------- 1 file changed, 26 insertions(+), 10 deletions(-) (limited to 'net') diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c index 6c1da21..9c66141 100644 --- a/net/netfilter/nf_conntrack_core.c +++ b/net/netfilter/nf_conntrack_core.c @@ -62,8 +62,8 @@ EXPORT_SYMBOL_GPL(nf_conntrack_htable_size); unsigned int nf_conntrack_max __read_mostly; EXPORT_SYMBOL_GPL(nf_conntrack_max); -struct nf_conn nf_conntrack_untracked; -EXPORT_SYMBOL_GPL(nf_conntrack_untracked); +DEFINE_PER_CPU(struct nf_conn, nf_conntrack_untracked); +EXPORT_PER_CPU_SYMBOL(nf_conntrack_untracked); static int nf_conntrack_hash_rnd_initted; static unsigned int nf_conntrack_hash_rnd; @@ -1183,10 +1183,21 @@ static void nf_ct_release_dying_list(struct net *net) spin_unlock_bh(&nf_conntrack_lock); } +static int untrack_refs(void) +{ + int cnt = 0, cpu; + + for_each_possible_cpu(cpu) { + struct nf_conn *ct = &per_cpu(nf_conntrack_untracked, cpu); + + cnt += atomic_read(&ct->ct_general.use) - 1; + } + return cnt; +} + static void nf_conntrack_cleanup_init_net(void) { - /* wait until all references to nf_conntrack_untracked are dropped */ - while (atomic_read(&nf_conntrack_untracked.ct_general.use) > 1) + while (untrack_refs() > 0) schedule(); nf_conntrack_helper_fini(); @@ -1323,14 +1334,17 @@ module_param_call(hashsize, nf_conntrack_set_hashsize, param_get_uint, void nf_ct_untracked_status_or(unsigned long bits) { - nf_conntrack_untracked.status |= bits; + int cpu; + + for_each_possible_cpu(cpu) + per_cpu(nf_conntrack_untracked, cpu).status |= bits; } EXPORT_SYMBOL_GPL(nf_ct_untracked_status_or); static int nf_conntrack_init_init_net(void) { int max_factor = 8; - int ret; + int ret, cpu; /* Idea from tcp.c: use 1/16384 of memory. On i386: 32MB * machine has 512 buckets. >= 1GB machines have 16384 buckets. */ @@ -1369,10 +1383,12 @@ static int nf_conntrack_init_init_net(void) goto err_extend; #endif /* Set up fake conntrack: to never be deleted, not in any hashes */ -#ifdef CONFIG_NET_NS - nf_conntrack_untracked.ct_net = &init_net; -#endif - atomic_set(&nf_conntrack_untracked.ct_general.use, 1); + for_each_possible_cpu(cpu) { + struct nf_conn *ct = &per_cpu(nf_conntrack_untracked, cpu); + + write_pnet(&ct->ct_net, &init_net); + atomic_set(&ct->ct_general.use, 1); + } /* - and look it like as a confirmed connection */ nf_ct_untracked_status_or(IPS_CONFIRMED | IPS_UNTRACKED); return 0; -- cgit v1.1 From 2bf074825403e0e0d623bac7573580773b78abef Mon Sep 17 00:00:00 2001 From: Shan Wei Date: Wed, 9 Jun 2010 14:47:40 +0200 Subject: netfilter: xt_sctp: use WORD_ROUND macro to calculate length of multiple of 4 bytes Use WORD_ROUND to round an int up to the next multiple of 4. Signed-off-by: Shan Wei Signed-off-by: Patrick McHardy --- net/netfilter/xt_sctp.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'net') diff --git a/net/netfilter/xt_sctp.c b/net/netfilter/xt_sctp.c index c04fcf3..ef36a56 100644 --- a/net/netfilter/xt_sctp.c +++ b/net/netfilter/xt_sctp.c @@ -3,6 +3,7 @@ #include #include #include +#include #include #include @@ -67,7 +68,7 @@ match_packet(const struct sk_buff *skb, ++i, offset, sch->type, htons(sch->length), sch->flags); #endif - offset += (ntohs(sch->length) + 3) & ~3; + offset += WORD_ROUND(ntohs(sch->length)); pr_debug("skb->len: %d\toffset: %d\n", skb->len, offset); -- cgit v1.1 From 5756d346c7cdefcd84a8ac4901167cdfb5447b69 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 9 Jun 2010 15:47:41 +0200 Subject: netfilter: ip_queue: rwlock to spinlock conversion Converts queue_lock rwlock to a spinlock. (readlocked part can be changed by reads of integer values) One atomic operation instead of four per ipq_enqueue_packet() call. Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/ipv4/netfilter/ip_queue.c | 57 +++++++++++++++++++------------------------ 1 file changed, 25 insertions(+), 32 deletions(-) (limited to 'net') diff --git a/net/ipv4/netfilter/ip_queue.c b/net/ipv4/netfilter/ip_queue.c index a4e5fc5..d2c1311 100644 --- a/net/ipv4/netfilter/ip_queue.c +++ b/net/ipv4/netfilter/ip_queue.c @@ -42,7 +42,7 @@ typedef int (*ipq_cmpfn)(struct nf_queue_entry *, unsigned long); static unsigned char copy_mode __read_mostly = IPQ_COPY_NONE; static unsigned int queue_maxlen __read_mostly = IPQ_QMAX_DEFAULT; -static DEFINE_RWLOCK(queue_lock); +static DEFINE_SPINLOCK(queue_lock); static int peer_pid __read_mostly; static unsigned int copy_range __read_mostly; static unsigned int queue_total; @@ -72,10 +72,10 @@ __ipq_set_mode(unsigned char mode, unsigned int range) break; case IPQ_COPY_PACKET: - copy_mode = mode; + if (range > 0xFFFF) + range = 0xFFFF; copy_range = range; - if (copy_range > 0xFFFF) - copy_range = 0xFFFF; + copy_mode = mode; break; default: @@ -101,7 +101,7 @@ ipq_find_dequeue_entry(unsigned long id) { struct nf_queue_entry *entry = NULL, *i; - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); list_for_each_entry(i, &queue_list, list) { if ((unsigned long)i == id) { @@ -115,7 +115,7 @@ ipq_find_dequeue_entry(unsigned long id) queue_total--; } - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return entry; } @@ -136,9 +136,9 @@ __ipq_flush(ipq_cmpfn cmpfn, unsigned long data) static void ipq_flush(ipq_cmpfn cmpfn, unsigned long data) { - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); __ipq_flush(cmpfn, data); - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); } static struct sk_buff * @@ -152,9 +152,7 @@ ipq_build_packet_message(struct nf_queue_entry *entry, int *errp) struct nlmsghdr *nlh; struct timeval tv; - read_lock_bh(&queue_lock); - - switch (copy_mode) { + switch (ACCESS_ONCE(copy_mode)) { case IPQ_COPY_META: case IPQ_COPY_NONE: size = NLMSG_SPACE(sizeof(*pmsg)); @@ -162,26 +160,21 @@ ipq_build_packet_message(struct nf_queue_entry *entry, int *errp) case IPQ_COPY_PACKET: if (entry->skb->ip_summed == CHECKSUM_PARTIAL && - (*errp = skb_checksum_help(entry->skb))) { - read_unlock_bh(&queue_lock); + (*errp = skb_checksum_help(entry->skb))) return NULL; - } - if (copy_range == 0 || copy_range > entry->skb->len) + + data_len = ACCESS_ONCE(copy_range); + if (data_len == 0 || data_len > entry->skb->len) data_len = entry->skb->len; - else - data_len = copy_range; size = NLMSG_SPACE(sizeof(*pmsg) + data_len); break; default: *errp = -EINVAL; - read_unlock_bh(&queue_lock); return NULL; } - read_unlock_bh(&queue_lock); - skb = alloc_skb(size, GFP_ATOMIC); if (!skb) goto nlmsg_failure; @@ -242,7 +235,7 @@ ipq_enqueue_packet(struct nf_queue_entry *entry, unsigned int queuenum) if (nskb == NULL) return status; - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); if (!peer_pid) goto err_out_free_nskb; @@ -266,14 +259,14 @@ ipq_enqueue_packet(struct nf_queue_entry *entry, unsigned int queuenum) __ipq_enqueue_entry(entry); - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return status; err_out_free_nskb: kfree_skb(nskb); err_out_unlock: - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return status; } @@ -342,9 +335,9 @@ ipq_set_mode(unsigned char mode, unsigned int range) { int status; - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); status = __ipq_set_mode(mode, range); - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return status; } @@ -440,11 +433,11 @@ __ipq_rcv_skb(struct sk_buff *skb) if (security_netlink_recv(skb, CAP_NET_ADMIN)) RCV_SKB_FAIL(-EPERM); - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); if (peer_pid) { if (peer_pid != pid) { - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); RCV_SKB_FAIL(-EBUSY); } } else { @@ -452,7 +445,7 @@ __ipq_rcv_skb(struct sk_buff *skb) peer_pid = pid; } - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); status = ipq_receive_peer(NLMSG_DATA(nlh), type, nlmsglen - NLMSG_LENGTH(0)); @@ -497,10 +490,10 @@ ipq_rcv_nl_event(struct notifier_block *this, struct netlink_notify *n = ptr; if (event == NETLINK_URELEASE && n->protocol == NETLINK_FIREWALL) { - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); if ((net_eq(n->net, &init_net)) && (n->pid == peer_pid)) __ipq_reset(); - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); } return NOTIFY_DONE; } @@ -527,7 +520,7 @@ static ctl_table ipq_table[] = { #ifdef CONFIG_PROC_FS static int ip_queue_show(struct seq_file *m, void *v) { - read_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); seq_printf(m, "Peer PID : %d\n" @@ -545,7 +538,7 @@ static int ip_queue_show(struct seq_file *m, void *v) queue_dropped, queue_user_dropped); - read_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return 0; } -- cgit v1.1 From 144ad2a6c56b6109ff0f64074863ae5cf1c1698a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 9 Jun 2010 16:25:08 +0200 Subject: netfilter: ip6_queue: rwlock to spinlock conversion Converts queue_lock rwlock to a spinlock. (readlocked part can be changed by reads of integer values) One atomic operation instead of four per ipq_enqueue_packet() call. Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/ipv6/netfilter/ip6_queue.c | 57 ++++++++++++++++++------------------------ 1 file changed, 25 insertions(+), 32 deletions(-) (limited to 'net') diff --git a/net/ipv6/netfilter/ip6_queue.c b/net/ipv6/netfilter/ip6_queue.c index 8c20174..413ab07 100644 --- a/net/ipv6/netfilter/ip6_queue.c +++ b/net/ipv6/netfilter/ip6_queue.c @@ -43,7 +43,7 @@ typedef int (*ipq_cmpfn)(struct nf_queue_entry *, unsigned long); static unsigned char copy_mode __read_mostly = IPQ_COPY_NONE; static unsigned int queue_maxlen __read_mostly = IPQ_QMAX_DEFAULT; -static DEFINE_RWLOCK(queue_lock); +static DEFINE_SPINLOCK(queue_lock); static int peer_pid __read_mostly; static unsigned int copy_range __read_mostly; static unsigned int queue_total; @@ -73,10 +73,10 @@ __ipq_set_mode(unsigned char mode, unsigned int range) break; case IPQ_COPY_PACKET: - copy_mode = mode; + if (range > 0xFFFF) + range = 0xFFFF; copy_range = range; - if (copy_range > 0xFFFF) - copy_range = 0xFFFF; + copy_mode = mode; break; default: @@ -102,7 +102,7 @@ ipq_find_dequeue_entry(unsigned long id) { struct nf_queue_entry *entry = NULL, *i; - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); list_for_each_entry(i, &queue_list, list) { if ((unsigned long)i == id) { @@ -116,7 +116,7 @@ ipq_find_dequeue_entry(unsigned long id) queue_total--; } - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return entry; } @@ -137,9 +137,9 @@ __ipq_flush(ipq_cmpfn cmpfn, unsigned long data) static void ipq_flush(ipq_cmpfn cmpfn, unsigned long data) { - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); __ipq_flush(cmpfn, data); - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); } static struct sk_buff * @@ -153,9 +153,7 @@ ipq_build_packet_message(struct nf_queue_entry *entry, int *errp) struct nlmsghdr *nlh; struct timeval tv; - read_lock_bh(&queue_lock); - - switch (copy_mode) { + switch (ACCESS_ONCE(copy_mode)) { case IPQ_COPY_META: case IPQ_COPY_NONE: size = NLMSG_SPACE(sizeof(*pmsg)); @@ -163,26 +161,21 @@ ipq_build_packet_message(struct nf_queue_entry *entry, int *errp) case IPQ_COPY_PACKET: if (entry->skb->ip_summed == CHECKSUM_PARTIAL && - (*errp = skb_checksum_help(entry->skb))) { - read_unlock_bh(&queue_lock); + (*errp = skb_checksum_help(entry->skb))) return NULL; - } - if (copy_range == 0 || copy_range > entry->skb->len) + + data_len = ACCESS_ONCE(copy_range); + if (data_len == 0 || data_len > entry->skb->len) data_len = entry->skb->len; - else - data_len = copy_range; size = NLMSG_SPACE(sizeof(*pmsg) + data_len); break; default: *errp = -EINVAL; - read_unlock_bh(&queue_lock); return NULL; } - read_unlock_bh(&queue_lock); - skb = alloc_skb(size, GFP_ATOMIC); if (!skb) goto nlmsg_failure; @@ -242,7 +235,7 @@ ipq_enqueue_packet(struct nf_queue_entry *entry, unsigned int queuenum) if (nskb == NULL) return status; - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); if (!peer_pid) goto err_out_free_nskb; @@ -266,14 +259,14 @@ ipq_enqueue_packet(struct nf_queue_entry *entry, unsigned int queuenum) __ipq_enqueue_entry(entry); - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return status; err_out_free_nskb: kfree_skb(nskb); err_out_unlock: - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return status; } @@ -342,9 +335,9 @@ ipq_set_mode(unsigned char mode, unsigned int range) { int status; - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); status = __ipq_set_mode(mode, range); - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return status; } @@ -441,11 +434,11 @@ __ipq_rcv_skb(struct sk_buff *skb) if (security_netlink_recv(skb, CAP_NET_ADMIN)) RCV_SKB_FAIL(-EPERM); - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); if (peer_pid) { if (peer_pid != pid) { - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); RCV_SKB_FAIL(-EBUSY); } } else { @@ -453,7 +446,7 @@ __ipq_rcv_skb(struct sk_buff *skb) peer_pid = pid; } - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); status = ipq_receive_peer(NLMSG_DATA(nlh), type, nlmsglen - NLMSG_LENGTH(0)); @@ -498,10 +491,10 @@ ipq_rcv_nl_event(struct notifier_block *this, struct netlink_notify *n = ptr; if (event == NETLINK_URELEASE && n->protocol == NETLINK_IP6_FW) { - write_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); if ((net_eq(n->net, &init_net)) && (n->pid == peer_pid)) __ipq_reset(); - write_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); } return NOTIFY_DONE; } @@ -528,7 +521,7 @@ static ctl_table ipq_table[] = { #ifdef CONFIG_PROC_FS static int ip6_queue_show(struct seq_file *m, void *v) { - read_lock_bh(&queue_lock); + spin_lock_bh(&queue_lock); seq_printf(m, "Peer PID : %d\n" @@ -546,7 +539,7 @@ static int ip6_queue_show(struct seq_file *m, void *v) queue_dropped, queue_user_dropped); - read_unlock_bh(&queue_lock); + spin_unlock_bh(&queue_lock); return 0; } -- cgit v1.1 From c463ac972315a0c86bb20b8d35225baa75caf899 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 9 Jun 2010 18:07:06 +0200 Subject: netfilter: nfnetlink_queue: some optimizations - Use an atomic_t for id_sequence to avoid a spin_lock/spin_unlock pair - Group highly modified struct nfqnl_instance fields together Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/netfilter/nfnetlink_queue.c | 33 ++++++++++++++------------------- 1 file changed, 14 insertions(+), 19 deletions(-) (limited to 'net') diff --git a/net/netfilter/nfnetlink_queue.c b/net/netfilter/nfnetlink_queue.c index 12e1ab3..d05605b 100644 --- a/net/netfilter/nfnetlink_queue.c +++ b/net/netfilter/nfnetlink_queue.c @@ -46,17 +46,19 @@ struct nfqnl_instance { int peer_pid; unsigned int queue_maxlen; unsigned int copy_range; - unsigned int queue_total; unsigned int queue_dropped; unsigned int queue_user_dropped; - unsigned int id_sequence; /* 'sequence' of pkt ids */ u_int16_t queue_num; /* number of this queue */ u_int8_t copy_mode; - - spinlock_t lock; - +/* + * Following fields are dirtied for each queued packet, + * keep them in same cache line if possible. + */ + spinlock_t lock; + unsigned int queue_total; + atomic_t id_sequence; /* 'sequence' of pkt ids */ struct list_head queue_list; /* packets in queue */ }; @@ -238,32 +240,24 @@ nfqnl_build_packet_message(struct nfqnl_instance *queue, outdev = entry->outdev; - spin_lock_bh(&queue->lock); - - switch ((enum nfqnl_config_mode)queue->copy_mode) { + switch ((enum nfqnl_config_mode)ACCESS_ONCE(queue->copy_mode)) { case NFQNL_COPY_META: case NFQNL_COPY_NONE: break; case NFQNL_COPY_PACKET: if (entskb->ip_summed == CHECKSUM_PARTIAL && - skb_checksum_help(entskb)) { - spin_unlock_bh(&queue->lock); + skb_checksum_help(entskb)) return NULL; - } - if (queue->copy_range == 0 - || queue->copy_range > entskb->len) + + data_len = ACCESS_ONCE(queue->copy_range); + if (data_len == 0 || data_len > entskb->len) data_len = entskb->len; - else - data_len = queue->copy_range; size += nla_total_size(data_len); break; } - entry->id = queue->id_sequence++; - - spin_unlock_bh(&queue->lock); skb = alloc_skb(size, GFP_ATOMIC); if (!skb) @@ -278,6 +272,7 @@ nfqnl_build_packet_message(struct nfqnl_instance *queue, nfmsg->version = NFNETLINK_V0; nfmsg->res_id = htons(queue->queue_num); + entry->id = atomic_inc_return(&queue->id_sequence); pmsg.packet_id = htonl(entry->id); pmsg.hw_protocol = entskb->protocol; pmsg.hook = entry->hook; @@ -866,7 +861,7 @@ static int seq_show(struct seq_file *s, void *v) inst->peer_pid, inst->queue_total, inst->copy_mode, inst->copy_range, inst->queue_dropped, inst->queue_user_dropped, - inst->id_sequence, 1); + atomic_read(&inst->id_sequence), 1); } static const struct seq_operations nfqnl_seq_ops = { -- cgit v1.1 From bed1be20867d17a3eb2fb5e1613ebdc50c83b8aa Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 9 Jun 2010 18:14:58 +0200 Subject: netfilter: nfnetlink_log: RCU conversion - instances_lock becomes a spinlock - lockless lookups While nfnetlink_log probably not performance critical, using less rwlocks in our code is always welcomed... Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/netfilter/nfnetlink_log.c | 49 ++++++++++++++++++++++++------------------- 1 file changed, 27 insertions(+), 22 deletions(-) (limited to 'net') diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c index fc9a211..8ec23ec 100644 --- a/net/netfilter/nfnetlink_log.c +++ b/net/netfilter/nfnetlink_log.c @@ -66,9 +66,10 @@ struct nfulnl_instance { u_int16_t group_num; /* number of this queue */ u_int16_t flags; u_int8_t copy_mode; + struct rcu_head rcu; }; -static DEFINE_RWLOCK(instances_lock); +static DEFINE_SPINLOCK(instances_lock); static atomic_t global_seq; #define INSTANCE_BUCKETS 16 @@ -88,7 +89,7 @@ __instance_lookup(u_int16_t group_num) struct nfulnl_instance *inst; head = &instance_table[instance_hashfn(group_num)]; - hlist_for_each_entry(inst, pos, head, hlist) { + hlist_for_each_entry_rcu(inst, pos, head, hlist) { if (inst->group_num == group_num) return inst; } @@ -106,22 +107,26 @@ instance_lookup_get(u_int16_t group_num) { struct nfulnl_instance *inst; - read_lock_bh(&instances_lock); + rcu_read_lock_bh(); inst = __instance_lookup(group_num); if (inst) instance_get(inst); - read_unlock_bh(&instances_lock); + rcu_read_unlock_bh(); return inst; } +static void nfulnl_instance_free_rcu(struct rcu_head *head) +{ + kfree(container_of(head, struct nfulnl_instance, rcu)); + module_put(THIS_MODULE); +} + static void instance_put(struct nfulnl_instance *inst) { - if (inst && atomic_dec_and_test(&inst->use)) { - kfree(inst); - module_put(THIS_MODULE); - } + if (inst && atomic_dec_and_test(&inst->use)) + call_rcu_bh(&inst->rcu, nfulnl_instance_free_rcu); } static void nfulnl_timer(unsigned long data); @@ -132,7 +137,7 @@ instance_create(u_int16_t group_num, int pid) struct nfulnl_instance *inst; int err; - write_lock_bh(&instances_lock); + spin_lock_bh(&instances_lock); if (__instance_lookup(group_num)) { err = -EEXIST; goto out_unlock; @@ -169,12 +174,12 @@ instance_create(u_int16_t group_num, int pid) hlist_add_head(&inst->hlist, &instance_table[instance_hashfn(group_num)]); - write_unlock_bh(&instances_lock); + spin_unlock_bh(&instances_lock); return inst; out_unlock: - write_unlock_bh(&instances_lock); + spin_unlock_bh(&instances_lock); return ERR_PTR(err); } @@ -200,9 +205,9 @@ __instance_destroy(struct nfulnl_instance *inst) static inline void instance_destroy(struct nfulnl_instance *inst) { - write_lock_bh(&instances_lock); + spin_lock_bh(&instances_lock); __instance_destroy(inst); - write_unlock_bh(&instances_lock); + spin_unlock_bh(&instances_lock); } static int @@ -672,7 +677,7 @@ nfulnl_rcv_nl_event(struct notifier_block *this, int i; /* destroy all instances for this pid */ - write_lock_bh(&instances_lock); + spin_lock_bh(&instances_lock); for (i = 0; i < INSTANCE_BUCKETS; i++) { struct hlist_node *tmp, *t2; struct nfulnl_instance *inst; @@ -684,7 +689,7 @@ nfulnl_rcv_nl_event(struct notifier_block *this, __instance_destroy(inst); } } - write_unlock_bh(&instances_lock); + spin_unlock_bh(&instances_lock); } return NOTIFY_DONE; } @@ -861,19 +866,19 @@ static struct hlist_node *get_first(struct iter_state *st) for (st->bucket = 0; st->bucket < INSTANCE_BUCKETS; st->bucket++) { if (!hlist_empty(&instance_table[st->bucket])) - return instance_table[st->bucket].first; + return rcu_dereference_bh(instance_table[st->bucket].first); } return NULL; } static struct hlist_node *get_next(struct iter_state *st, struct hlist_node *h) { - h = h->next; + h = rcu_dereference_bh(h->next); while (!h) { if (++st->bucket >= INSTANCE_BUCKETS) return NULL; - h = instance_table[st->bucket].first; + h = rcu_dereference_bh(instance_table[st->bucket].first); } return h; } @@ -890,9 +895,9 @@ static struct hlist_node *get_idx(struct iter_state *st, loff_t pos) } static void *seq_start(struct seq_file *seq, loff_t *pos) - __acquires(instances_lock) + __acquires(rcu_bh) { - read_lock_bh(&instances_lock); + rcu_read_lock_bh(); return get_idx(seq->private, *pos); } @@ -903,9 +908,9 @@ static void *seq_next(struct seq_file *s, void *v, loff_t *pos) } static void seq_stop(struct seq_file *s, void *v) - __releases(instances_lock) + __releases(rcu_bh) { - read_unlock_bh(&instances_lock); + rcu_read_unlock_bh(); } static int seq_show(struct seq_file *s, void *v) -- cgit v1.1 From 88e7594a9775e54dcd421cb246406dce62e48bee Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Mon, 7 Jun 2010 03:27:39 +0000 Subject: phonet: use call_rcu for phonet device free Use call_rcu rather than synchronize_rcu. Signed-off-by: Jiri Pirko Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/phonet/pn_dev.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/phonet/pn_dev.c b/net/phonet/pn_dev.c index c33da65..b18e48f 100644 --- a/net/phonet/pn_dev.c +++ b/net/phonet/pn_dev.c @@ -162,6 +162,14 @@ int phonet_address_add(struct net_device *dev, u8 addr) return err; } +static void phonet_device_rcu_free(struct rcu_head *head) +{ + struct phonet_device *pnd; + + pnd = container_of(head, struct phonet_device, rcu); + kfree(pnd); +} + int phonet_address_del(struct net_device *dev, u8 addr) { struct phonet_device_list *pndevs = phonet_device_list(dev_net(dev)); @@ -179,10 +187,9 @@ int phonet_address_del(struct net_device *dev, u8 addr) pnd = NULL; mutex_unlock(&pndevs->lock); - if (pnd) { - synchronize_rcu(); - kfree(pnd); - } + if (pnd) + call_rcu(&pnd->rcu, phonet_device_rcu_free); + return err; } -- cgit v1.1 From cfa087f689402438e3cb0f077e649d01c871b0e7 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 7 Jun 2010 22:34:35 +0000 Subject: icmp: RCU conversion in icmp_address_reply() - rcu_read_lock() already held by caller - use __in_dev_get_rcu() instead of in_dev_get() / in_dev_put() - remove goto out; Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/icmp.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) (limited to 'net') diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index d65e9215..bdb6c71 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -925,6 +925,7 @@ static void icmp_address(struct sk_buff *skb) /* * RFC1812 (4.3.3.9). A router SHOULD listen all replies, and complain * loudly if an inconsistency is found. + * called with rcu_read_lock() */ static void icmp_address_reply(struct sk_buff *skb) @@ -935,12 +936,12 @@ static void icmp_address_reply(struct sk_buff *skb) struct in_ifaddr *ifa; if (skb->len < 4 || !(rt->rt_flags&RTCF_DIRECTSRC)) - goto out; + return; - in_dev = in_dev_get(dev); + in_dev = __in_dev_get_rcu(dev); if (!in_dev) - goto out; - rcu_read_lock(); + return; + if (in_dev->ifa_list && IN_DEV_LOG_MARTIANS(in_dev) && IN_DEV_FORWARD(in_dev)) { @@ -958,9 +959,6 @@ static void icmp_address_reply(struct sk_buff *skb) mp, dev->name, &rt->rt_src); } } - rcu_read_unlock(); - in_dev_put(in_dev); -out:; } static void icmp_discard(struct sk_buff *skb) -- cgit v1.1 From 96b52e61be1ad4d4f8de39b9deaf253da804ea3b Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 7 Jun 2010 21:05:02 +0000 Subject: ipv6: mcast: RCU conversions - ipv6_sock_mc_join() : doesnt touch dev refcount - ipv6_sock_mc_drop() : doesnt touch dev/idev refcounts - ip6_mc_find_dev() becomes ip6_mc_find_dev_rcu() (called from rcu), and doesnt touch dev/idev refcounts - ipv6_sock_mc_close() : doesnt touch dev/idev refcounts - ip6_mc_source() uses ip6_mc_find_dev_rcu() - ip6_mc_msfilter() uses ip6_mc_find_dev_rcu() - ip6_mc_msfget() uses ip6_mc_find_dev_rcu() - ipv6_dev_mc_dec(), ipv6_chk_mcast_addr(), igmp6_event_query(), igmp6_event_report(), mld_sendpack(), igmp6_send() dont touch idev refcount Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv6/mcast.c | 183 ++++++++++++++++++++++++++----------------------------- 1 file changed, 87 insertions(+), 96 deletions(-) (limited to 'net') diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c index 8752e80..3e36d15 100644 --- a/net/ipv6/mcast.c +++ b/net/ipv6/mcast.c @@ -152,18 +152,19 @@ int ipv6_sock_mc_join(struct sock *sk, int ifindex, const struct in6_addr *addr) mc_lst->next = NULL; ipv6_addr_copy(&mc_lst->addr, addr); + rcu_read_lock(); if (ifindex == 0) { struct rt6_info *rt; rt = rt6_lookup(net, addr, NULL, 0, 0); if (rt) { dev = rt->rt6i_dev; - dev_hold(dev); dst_release(&rt->u.dst); } } else - dev = dev_get_by_index(net, ifindex); + dev = dev_get_by_index_rcu(net, ifindex); if (dev == NULL) { + rcu_read_unlock(); sock_kfree_s(sk, mc_lst, sizeof(*mc_lst)); return -ENODEV; } @@ -180,8 +181,8 @@ int ipv6_sock_mc_join(struct sock *sk, int ifindex, const struct in6_addr *addr) err = ipv6_dev_mc_inc(dev, addr); if (err) { + rcu_read_unlock(); sock_kfree_s(sk, mc_lst, sizeof(*mc_lst)); - dev_put(dev); return err; } @@ -190,7 +191,7 @@ int ipv6_sock_mc_join(struct sock *sk, int ifindex, const struct in6_addr *addr) np->ipv6_mc_list = mc_lst; write_unlock_bh(&ipv6_sk_mc_lock); - dev_put(dev); + rcu_read_unlock(); return 0; } @@ -213,18 +214,17 @@ int ipv6_sock_mc_drop(struct sock *sk, int ifindex, const struct in6_addr *addr) *lnk = mc_lst->next; write_unlock_bh(&ipv6_sk_mc_lock); - dev = dev_get_by_index(net, mc_lst->ifindex); + rcu_read_lock(); + dev = dev_get_by_index_rcu(net, mc_lst->ifindex); if (dev != NULL) { - struct inet6_dev *idev = in6_dev_get(dev); + struct inet6_dev *idev = __in6_dev_get(dev); (void) ip6_mc_leave_src(sk, mc_lst, idev); - if (idev) { + if (idev) __ipv6_dev_mc_dec(idev, &mc_lst->addr); - in6_dev_put(idev); - } - dev_put(dev); } else (void) ip6_mc_leave_src(sk, mc_lst, NULL); + rcu_read_unlock(); sock_kfree_s(sk, mc_lst, sizeof(*mc_lst)); return 0; } @@ -234,43 +234,36 @@ int ipv6_sock_mc_drop(struct sock *sk, int ifindex, const struct in6_addr *addr) return -EADDRNOTAVAIL; } -static struct inet6_dev *ip6_mc_find_dev(struct net *net, - struct in6_addr *group, - int ifindex) +/* called with rcu_read_lock() */ +static struct inet6_dev *ip6_mc_find_dev_rcu(struct net *net, + struct in6_addr *group, + int ifindex) { struct net_device *dev = NULL; struct inet6_dev *idev = NULL; if (ifindex == 0) { - struct rt6_info *rt; + struct rt6_info *rt = rt6_lookup(net, group, NULL, 0, 0); - rt = rt6_lookup(net, group, NULL, 0, 0); if (rt) { dev = rt->rt6i_dev; dev_hold(dev); dst_release(&rt->u.dst); } } else - dev = dev_get_by_index(net, ifindex); + dev = dev_get_by_index_rcu(net, ifindex); if (!dev) - goto nodev; - idev = in6_dev_get(dev); + return NULL; + idev = __in6_dev_get(dev); if (!idev) - goto release; + return NULL;; read_lock_bh(&idev->lock); - if (idev->dead) - goto unlock_release; - + if (idev->dead) { + read_unlock_bh(&idev->lock); + return NULL; + } return idev; - -unlock_release: - read_unlock_bh(&idev->lock); - in6_dev_put(idev); -release: - dev_put(dev); -nodev: - return NULL; } void ipv6_sock_mc_close(struct sock *sk) @@ -286,19 +279,17 @@ void ipv6_sock_mc_close(struct sock *sk) np->ipv6_mc_list = mc_lst->next; write_unlock_bh(&ipv6_sk_mc_lock); - dev = dev_get_by_index(net, mc_lst->ifindex); + rcu_read_lock(); + dev = dev_get_by_index_rcu(net, mc_lst->ifindex); if (dev) { - struct inet6_dev *idev = in6_dev_get(dev); + struct inet6_dev *idev = __in6_dev_get(dev); (void) ip6_mc_leave_src(sk, mc_lst, idev); - if (idev) { + if (idev) __ipv6_dev_mc_dec(idev, &mc_lst->addr); - in6_dev_put(idev); - } - dev_put(dev); } else (void) ip6_mc_leave_src(sk, mc_lst, NULL); - + rcu_read_unlock(); sock_kfree_s(sk, mc_lst, sizeof(*mc_lst)); write_lock_bh(&ipv6_sk_mc_lock); @@ -327,14 +318,17 @@ int ip6_mc_source(int add, int omode, struct sock *sk, if (!ipv6_addr_is_multicast(group)) return -EINVAL; - idev = ip6_mc_find_dev(net, group, pgsr->gsr_interface); - if (!idev) + rcu_read_lock(); + idev = ip6_mc_find_dev_rcu(net, group, pgsr->gsr_interface); + if (!idev) { + rcu_read_unlock(); return -ENODEV; + } dev = idev->dev; err = -EADDRNOTAVAIL; - read_lock_bh(&ipv6_sk_mc_lock); + read_lock(&ipv6_sk_mc_lock); for (pmc=inet6->ipv6_mc_list; pmc; pmc=pmc->next) { if (pgsr->gsr_interface && pmc->ifindex != pgsr->gsr_interface) continue; @@ -358,7 +352,7 @@ int ip6_mc_source(int add, int omode, struct sock *sk, pmc->sfmode = omode; } - write_lock_bh(&pmc->sflock); + write_lock(&pmc->sflock); pmclocked = 1; psl = pmc->sflist; @@ -433,11 +427,10 @@ int ip6_mc_source(int add, int omode, struct sock *sk, ip6_mc_add_src(idev, group, omode, 1, source, 1); done: if (pmclocked) - write_unlock_bh(&pmc->sflock); - read_unlock_bh(&ipv6_sk_mc_lock); + write_unlock(&pmc->sflock); + read_unlock(&ipv6_sk_mc_lock); read_unlock_bh(&idev->lock); - in6_dev_put(idev); - dev_put(dev); + rcu_read_unlock(); if (leavegroup) return ipv6_sock_mc_drop(sk, pgsr->gsr_interface, group); return err; @@ -463,14 +456,17 @@ int ip6_mc_msfilter(struct sock *sk, struct group_filter *gsf) gsf->gf_fmode != MCAST_EXCLUDE) return -EINVAL; - idev = ip6_mc_find_dev(net, group, gsf->gf_interface); + rcu_read_lock(); + idev = ip6_mc_find_dev_rcu(net, group, gsf->gf_interface); - if (!idev) + if (!idev) { + rcu_read_unlock(); return -ENODEV; + } dev = idev->dev; err = 0; - read_lock_bh(&ipv6_sk_mc_lock); + read_lock(&ipv6_sk_mc_lock); if (gsf->gf_fmode == MCAST_INCLUDE && gsf->gf_numsrc == 0) { leavegroup = 1; @@ -512,7 +508,7 @@ int ip6_mc_msfilter(struct sock *sk, struct group_filter *gsf) (void) ip6_mc_add_src(idev, group, gsf->gf_fmode, 0, NULL, 0); } - write_lock_bh(&pmc->sflock); + write_lock(&pmc->sflock); psl = pmc->sflist; if (psl) { (void) ip6_mc_del_src(idev, group, pmc->sfmode, @@ -522,13 +518,12 @@ int ip6_mc_msfilter(struct sock *sk, struct group_filter *gsf) (void) ip6_mc_del_src(idev, group, pmc->sfmode, 0, NULL, 0); pmc->sflist = newpsl; pmc->sfmode = gsf->gf_fmode; - write_unlock_bh(&pmc->sflock); + write_unlock(&pmc->sflock); err = 0; done: - read_unlock_bh(&ipv6_sk_mc_lock); + read_unlock(&ipv6_sk_mc_lock); read_unlock_bh(&idev->lock); - in6_dev_put(idev); - dev_put(dev); + rcu_read_unlock(); if (leavegroup) err = ipv6_sock_mc_drop(sk, gsf->gf_interface, group); return err; @@ -551,11 +546,13 @@ int ip6_mc_msfget(struct sock *sk, struct group_filter *gsf, if (!ipv6_addr_is_multicast(group)) return -EINVAL; - idev = ip6_mc_find_dev(net, group, gsf->gf_interface); + rcu_read_lock(); + idev = ip6_mc_find_dev_rcu(net, group, gsf->gf_interface); - if (!idev) + if (!idev) { + rcu_read_unlock(); return -ENODEV; - + } dev = idev->dev; err = -EADDRNOTAVAIL; @@ -577,8 +574,7 @@ int ip6_mc_msfget(struct sock *sk, struct group_filter *gsf, psl = pmc->sflist; count = psl ? psl->sl_count : 0; read_unlock_bh(&idev->lock); - in6_dev_put(idev); - dev_put(dev); + rcu_read_unlock(); copycount = count < gsf->gf_numsrc ? count : gsf->gf_numsrc; gsf->gf_numsrc = count; @@ -604,8 +600,7 @@ int ip6_mc_msfget(struct sock *sk, struct group_filter *gsf, return 0; done: read_unlock_bh(&idev->lock); - in6_dev_put(idev); - dev_put(dev); + rcu_read_unlock(); return err; } @@ -822,6 +817,7 @@ int ipv6_dev_mc_inc(struct net_device *dev, const struct in6_addr *addr) struct ifmcaddr6 *mc; struct inet6_dev *idev; + /* we need to take a reference on idev */ idev = in6_dev_get(dev); if (idev == NULL) @@ -860,7 +856,7 @@ int ipv6_dev_mc_inc(struct net_device *dev, const struct in6_addr *addr) setup_timer(&mc->mca_timer, igmp6_timer_handler, (unsigned long)mc); ipv6_addr_copy(&mc->mca_addr, addr); - mc->idev = idev; + mc->idev = idev; /* (reference taken) */ mc->mca_users = 1; /* mca_stamp should be updated upon changes */ mc->mca_cstamp = mc->mca_tstamp = jiffies; @@ -915,16 +911,18 @@ int __ipv6_dev_mc_dec(struct inet6_dev *idev, const struct in6_addr *addr) int ipv6_dev_mc_dec(struct net_device *dev, const struct in6_addr *addr) { - struct inet6_dev *idev = in6_dev_get(dev); + struct inet6_dev *idev; int err; - if (!idev) - return -ENODEV; - - err = __ipv6_dev_mc_dec(idev, addr); + rcu_read_lock(); - in6_dev_put(idev); + idev = __in6_dev_get(dev); + if (!idev) + err = -ENODEV; + else + err = __ipv6_dev_mc_dec(idev, addr); + rcu_read_unlock(); return err; } @@ -965,7 +963,8 @@ int ipv6_chk_mcast_addr(struct net_device *dev, const struct in6_addr *group, struct ifmcaddr6 *mc; int rv = 0; - idev = in6_dev_get(dev); + rcu_read_lock(); + idev = __in6_dev_get(dev); if (idev) { read_lock_bh(&idev->lock); for (mc = idev->mc_list; mc; mc=mc->next) { @@ -992,8 +991,8 @@ int ipv6_chk_mcast_addr(struct net_device *dev, const struct in6_addr *group, rv = 1; /* don't filter unspecified source */ } read_unlock_bh(&idev->lock); - in6_dev_put(idev); } + rcu_read_unlock(); return rv; } @@ -1104,6 +1103,7 @@ static int mld_marksources(struct ifmcaddr6 *pmc, int nsrcs, return 1; } +/* called with rcu_read_lock() */ int igmp6_event_query(struct sk_buff *skb) { struct mld2_query *mlh2 = NULL; @@ -1127,7 +1127,7 @@ int igmp6_event_query(struct sk_buff *skb) if (!(ipv6_addr_type(&ipv6_hdr(skb)->saddr) & IPV6_ADDR_LINKLOCAL)) return -EINVAL; - idev = in6_dev_get(skb->dev); + idev = __in6_dev_get(skb->dev); if (idev == NULL) return 0; @@ -1137,10 +1137,8 @@ int igmp6_event_query(struct sk_buff *skb) group_type = ipv6_addr_type(group); if (group_type != IPV6_ADDR_ANY && - !(group_type&IPV6_ADDR_MULTICAST)) { - in6_dev_put(idev); + !(group_type&IPV6_ADDR_MULTICAST)) return -EINVAL; - } if (len == 24) { int switchback; @@ -1161,10 +1159,9 @@ int igmp6_event_query(struct sk_buff *skb) } else if (len >= 28) { int srcs_offset = sizeof(struct mld2_query) - sizeof(struct icmp6hdr); - if (!pskb_may_pull(skb, srcs_offset)) { - in6_dev_put(idev); + if (!pskb_may_pull(skb, srcs_offset)) return -EINVAL; - } + mlh2 = (struct mld2_query *)skb_transport_header(skb); max_delay = (MLDV2_MRC(ntohs(mlh2->mld2q_mrc))*HZ)/1000; if (!max_delay) @@ -1173,28 +1170,23 @@ int igmp6_event_query(struct sk_buff *skb) if (mlh2->mld2q_qrv) idev->mc_qrv = mlh2->mld2q_qrv; if (group_type == IPV6_ADDR_ANY) { /* general query */ - if (mlh2->mld2q_nsrcs) { - in6_dev_put(idev); + if (mlh2->mld2q_nsrcs) return -EINVAL; /* no sources allowed */ - } + mld_gq_start_timer(idev); - in6_dev_put(idev); return 0; } /* mark sources to include, if group & source-specific */ if (mlh2->mld2q_nsrcs != 0) { if (!pskb_may_pull(skb, srcs_offset + - ntohs(mlh2->mld2q_nsrcs) * sizeof(struct in6_addr))) { - in6_dev_put(idev); + ntohs(mlh2->mld2q_nsrcs) * sizeof(struct in6_addr))) return -EINVAL; - } + mlh2 = (struct mld2_query *)skb_transport_header(skb); mark = 1; } - } else { - in6_dev_put(idev); + } else return -EINVAL; - } read_lock_bh(&idev->lock); if (group_type == IPV6_ADDR_ANY) { @@ -1227,12 +1219,11 @@ int igmp6_event_query(struct sk_buff *skb) } } read_unlock_bh(&idev->lock); - in6_dev_put(idev); return 0; } - +/* called with rcu_read_lock() */ int igmp6_event_report(struct sk_buff *skb) { struct ifmcaddr6 *ma; @@ -1260,7 +1251,7 @@ int igmp6_event_report(struct sk_buff *skb) !(addr_type&IPV6_ADDR_LINKLOCAL)) return -EINVAL; - idev = in6_dev_get(skb->dev); + idev = __in6_dev_get(skb->dev); if (idev == NULL) return -ENODEV; @@ -1280,7 +1271,6 @@ int igmp6_event_report(struct sk_buff *skb) } } read_unlock_bh(&idev->lock); - in6_dev_put(idev); return 0; } @@ -1396,12 +1386,14 @@ static void mld_sendpack(struct sk_buff *skb) struct mld2_report *pmr = (struct mld2_report *)skb_transport_header(skb); int payload_len, mldlen; - struct inet6_dev *idev = in6_dev_get(skb->dev); + struct inet6_dev *idev; struct net *net = dev_net(skb->dev); int err; struct flowi fl; struct dst_entry *dst; + rcu_read_lock(); + idev = __in6_dev_get(skb->dev); IP6_UPD_PO_STATS(net, idev, IPSTATS_MIB_OUT, skb->len); payload_len = (skb->tail - skb->network_header) - sizeof(*pip6); @@ -1441,8 +1433,7 @@ out: } else IP6_INC_STATS_BH(net, idev, IPSTATS_MIB_OUTDISCARDS); - if (likely(idev != NULL)) - in6_dev_put(idev); + rcu_read_unlock(); return; err_out: @@ -1779,7 +1770,8 @@ static void igmp6_send(struct in6_addr *addr, struct net_device *dev, int type) IPPROTO_ICMPV6, csum_partial(hdr, len, 0)); - idev = in6_dev_get(skb->dev); + rcu_read_lock(); + idev = __in6_dev_get(skb->dev); dst = icmp6_dst_alloc(skb->dev, NULL, &ipv6_hdr(skb)->daddr); if (!dst) { @@ -1806,8 +1798,7 @@ out: } else IP6_INC_STATS(net, idev, IPSTATS_MIB_OUTDISCARDS); - if (likely(idev != NULL)) - in6_dev_put(idev); + rcu_read_unlock(); return; err_out: -- cgit v1.1 From 592fcb9dfafaa02dd0edc207bf5d3a0ee7a1f8df Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 9 Jun 2010 16:21:07 +0000 Subject: ip: ip_ra_control() rcu fix commit 66018506e15b (ip: Router Alert RCU conversion) introduced RCU lookups to ip_call_ra_chain(). It missed proper deinit phase : When ip_ra_control() deletes an ip_ra_chain, it should make sure ip_call_ra_chain() users can not start to use socket during the rcu grace period. It should also delay the sock_put() after the grace period, or we risk a premature socket freeing and corruptions, as raw sockets are not rcu protected yet. This delay avoids using expensive atomic_inc_not_zero() in ip_call_ra_chain(). Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/ip_sockglue.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c index 08b9519..47fff52 100644 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@ -241,9 +241,13 @@ int ip_cmsg_send(struct net *net, struct msghdr *msg, struct ipcm_cookie *ipc) struct ip_ra_chain *ip_ra_chain; static DEFINE_SPINLOCK(ip_ra_lock); -static void ip_ra_free_rcu(struct rcu_head *head) + +static void ip_ra_destroy_rcu(struct rcu_head *head) { - kfree(container_of(head, struct ip_ra_chain, rcu)); + struct ip_ra_chain *ra = container_of(head, struct ip_ra_chain, rcu); + + sock_put(ra->saved_sk); + kfree(ra); } int ip_ra_control(struct sock *sk, unsigned char on, @@ -264,13 +268,20 @@ int ip_ra_control(struct sock *sk, unsigned char on, kfree(new_ra); return -EADDRINUSE; } + /* dont let ip_call_ra_chain() use sk again */ + ra->sk = NULL; rcu_assign_pointer(*rap, ra->next); spin_unlock_bh(&ip_ra_lock); if (ra->destructor) ra->destructor(sk); - sock_put(sk); - call_rcu(&ra->rcu, ip_ra_free_rcu); + /* + * Delay sock_put(sk) and kfree(ra) after one rcu grace + * period. This guarantee ip_call_ra_chain() dont need + * to mess with socket refcounts. + */ + ra->saved_sk = sk; + call_rcu(&ra->rcu, ip_ra_destroy_rcu); return 0; } } -- cgit v1.1 From d8d1f30b95a635dbd610dcc5eb641aca8f4768cf Mon Sep 17 00:00:00 2001 From: Changli Gao Date: Thu, 10 Jun 2010 23:31:35 -0700 Subject: net-next: remove useless union keyword remove useless union keyword in rtable, rt6_info and dn_route. Since there is only one member in a union, the union keyword isn't useful. Signed-off-by: Changli Gao Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/atm/clip.c | 2 +- net/bridge/br_device.c | 2 +- net/bridge/br_netfilter.c | 20 +- net/dccp/ipv4.c | 4 +- net/decnet/dn_route.c | 158 ++++++------ net/ethernet/eth.c | 5 +- net/ipv4/af_inet.c | 4 +- net/ipv4/arp.c | 12 +- net/ipv4/datagram.c | 2 +- net/ipv4/icmp.c | 18 +- net/ipv4/igmp.c | 10 +- net/ipv4/inet_connection_sock.c | 2 +- net/ipv4/ip_forward.c | 10 +- net/ipv4/ip_gre.c | 14 +- net/ipv4/ip_input.c | 4 +- net/ipv4/ip_output.c | 60 ++--- net/ipv4/ipip.c | 8 +- net/ipv4/ipmr.c | 8 +- net/ipv4/netfilter.c | 8 +- net/ipv4/raw.c | 16 +- net/ipv4/route.c | 420 ++++++++++++++++---------------- net/ipv4/syncookies.c | 6 +- net/ipv4/tcp_ipv4.c | 2 +- net/ipv4/udp.c | 4 +- net/ipv4/xfrm4_policy.c | 2 +- net/ipv6/addrconf.c | 10 +- net/ipv6/anycast.c | 6 +- net/ipv6/fib6_rules.c | 10 +- net/ipv6/ip6_fib.c | 30 +-- net/ipv6/ip6_output.c | 38 +-- net/ipv6/ip6_tunnel.c | 8 +- net/ipv6/mcast.c | 4 +- net/ipv6/ndisc.c | 8 +- net/ipv6/raw.c | 12 +- net/ipv6/route.c | 300 +++++++++++------------ net/ipv6/sit.c | 8 +- net/l2tp/l2tp_ip.c | 6 +- net/netfilter/ipvs/ip_vs_xmit.c | 86 +++---- net/netfilter/nf_conntrack_h323_main.c | 12 +- net/netfilter/nf_conntrack_netbios_ns.c | 2 +- net/netfilter/xt_TCPMSS.c | 4 +- net/netfilter/xt_TEE.c | 4 +- net/rxrpc/ar-peer.c | 4 +- net/sctp/protocol.c | 4 +- 44 files changed, 674 insertions(+), 683 deletions(-) (limited to 'net') diff --git a/net/atm/clip.c b/net/atm/clip.c index 313aba1..95fdd11 100644 --- a/net/atm/clip.c +++ b/net/atm/clip.c @@ -522,7 +522,7 @@ static int clip_setentry(struct atm_vcc *vcc, __be32 ip) error = ip_route_output_key(&init_net, &rt, &fl); if (error) return error; - neigh = __neigh_lookup(&clip_tbl, &ip, rt->u.dst.dev, 1); + neigh = __neigh_lookup(&clip_tbl, &ip, rt->dst.dev, 1); ip_rt_put(rt); if (!neigh) return -ENOMEM; diff --git a/net/bridge/br_device.c b/net/bridge/br_device.c index eedf2c9..b898364 100644 --- a/net/bridge/br_device.c +++ b/net/bridge/br_device.c @@ -127,7 +127,7 @@ static int br_change_mtu(struct net_device *dev, int new_mtu) #ifdef CONFIG_BRIDGE_NETFILTER /* remember the MTU in the rtable for PMTU */ - br->fake_rtable.u.dst.metrics[RTAX_MTU - 1] = new_mtu; + br->fake_rtable.dst.metrics[RTAX_MTU - 1] = new_mtu; #endif return 0; diff --git a/net/bridge/br_netfilter.c b/net/bridge/br_netfilter.c index 4442099..0685b25 100644 --- a/net/bridge/br_netfilter.c +++ b/net/bridge/br_netfilter.c @@ -117,12 +117,12 @@ void br_netfilter_rtable_init(struct net_bridge *br) { struct rtable *rt = &br->fake_rtable; - atomic_set(&rt->u.dst.__refcnt, 1); - rt->u.dst.dev = br->dev; - rt->u.dst.path = &rt->u.dst; - rt->u.dst.metrics[RTAX_MTU - 1] = 1500; - rt->u.dst.flags = DST_NOXFRM; - rt->u.dst.ops = &fake_dst_ops; + atomic_set(&rt->dst.__refcnt, 1); + rt->dst.dev = br->dev; + rt->dst.path = &rt->dst; + rt->dst.metrics[RTAX_MTU - 1] = 1500; + rt->dst.flags = DST_NOXFRM; + rt->dst.ops = &fake_dst_ops; } static inline struct rtable *bridge_parent_rtable(const struct net_device *dev) @@ -244,8 +244,8 @@ static int br_nf_pre_routing_finish_ipv6(struct sk_buff *skb) kfree_skb(skb); return 0; } - dst_hold(&rt->u.dst); - skb_dst_set(skb, &rt->u.dst); + dst_hold(&rt->dst); + skb_dst_set(skb, &rt->dst); skb->dev = nf_bridge->physindev; nf_bridge_update_protocol(skb); @@ -396,8 +396,8 @@ bridged_dnat: kfree_skb(skb); return 0; } - dst_hold(&rt->u.dst); - skb_dst_set(skb, &rt->u.dst); + dst_hold(&rt->dst); + skb_dst_set(skb, &rt->dst); } skb->dev = nf_bridge->physindev; diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c index d9b11ef..d4a166f 100644 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@ -105,7 +105,7 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) goto failure; /* OK, now commit destination to socket. */ - sk_setup_caps(sk, &rt->u.dst); + sk_setup_caps(sk, &rt->dst); dp->dccps_iss = secure_dccp_sequence_number(inet->inet_saddr, inet->inet_daddr, @@ -475,7 +475,7 @@ static struct dst_entry* dccp_v4_route_skb(struct net *net, struct sock *sk, return NULL; } - return &rt->u.dst; + return &rt->dst; } static int dccp_v4_send_response(struct sock *sk, struct request_sock *req, diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c index 812e6df..6585ea6 100644 --- a/net/decnet/dn_route.c +++ b/net/decnet/dn_route.c @@ -146,13 +146,13 @@ static __inline__ unsigned dn_hash(__le16 src, __le16 dst) static inline void dnrt_free(struct dn_route *rt) { - call_rcu_bh(&rt->u.dst.rcu_head, dst_rcu_free); + call_rcu_bh(&rt->dst.rcu_head, dst_rcu_free); } static inline void dnrt_drop(struct dn_route *rt) { - dst_release(&rt->u.dst); - call_rcu_bh(&rt->u.dst.rcu_head, dst_rcu_free); + dst_release(&rt->dst); + call_rcu_bh(&rt->dst.rcu_head, dst_rcu_free); } static void dn_dst_check_expire(unsigned long dummy) @@ -167,13 +167,13 @@ static void dn_dst_check_expire(unsigned long dummy) spin_lock(&dn_rt_hash_table[i].lock); while((rt=*rtp) != NULL) { - if (atomic_read(&rt->u.dst.__refcnt) || - (now - rt->u.dst.lastuse) < expire) { - rtp = &rt->u.dst.dn_next; + if (atomic_read(&rt->dst.__refcnt) || + (now - rt->dst.lastuse) < expire) { + rtp = &rt->dst.dn_next; continue; } - *rtp = rt->u.dst.dn_next; - rt->u.dst.dn_next = NULL; + *rtp = rt->dst.dn_next; + rt->dst.dn_next = NULL; dnrt_free(rt); } spin_unlock(&dn_rt_hash_table[i].lock); @@ -198,13 +198,13 @@ static int dn_dst_gc(struct dst_ops *ops) rtp = &dn_rt_hash_table[i].chain; while((rt=*rtp) != NULL) { - if (atomic_read(&rt->u.dst.__refcnt) || - (now - rt->u.dst.lastuse) < expire) { - rtp = &rt->u.dst.dn_next; + if (atomic_read(&rt->dst.__refcnt) || + (now - rt->dst.lastuse) < expire) { + rtp = &rt->dst.dn_next; continue; } - *rtp = rt->u.dst.dn_next; - rt->u.dst.dn_next = NULL; + *rtp = rt->dst.dn_next; + rt->dst.dn_next = NULL; dnrt_drop(rt); break; } @@ -287,25 +287,25 @@ static int dn_insert_route(struct dn_route *rt, unsigned hash, struct dn_route * while((rth = *rthp) != NULL) { if (compare_keys(&rth->fl, &rt->fl)) { /* Put it first */ - *rthp = rth->u.dst.dn_next; - rcu_assign_pointer(rth->u.dst.dn_next, + *rthp = rth->dst.dn_next; + rcu_assign_pointer(rth->dst.dn_next, dn_rt_hash_table[hash].chain); rcu_assign_pointer(dn_rt_hash_table[hash].chain, rth); - dst_use(&rth->u.dst, now); + dst_use(&rth->dst, now); spin_unlock_bh(&dn_rt_hash_table[hash].lock); dnrt_drop(rt); *rp = rth; return 0; } - rthp = &rth->u.dst.dn_next; + rthp = &rth->dst.dn_next; } - rcu_assign_pointer(rt->u.dst.dn_next, dn_rt_hash_table[hash].chain); + rcu_assign_pointer(rt->dst.dn_next, dn_rt_hash_table[hash].chain); rcu_assign_pointer(dn_rt_hash_table[hash].chain, rt); - dst_use(&rt->u.dst, now); + dst_use(&rt->dst, now); spin_unlock_bh(&dn_rt_hash_table[hash].lock); *rp = rt; return 0; @@ -323,8 +323,8 @@ static void dn_run_flush(unsigned long dummy) goto nothing_to_declare; for(; rt; rt=next) { - next = rt->u.dst.dn_next; - rt->u.dst.dn_next = NULL; + next = rt->dst.dn_next; + rt->dst.dn_next = NULL; dst_free((struct dst_entry *)rt); } @@ -743,7 +743,7 @@ static int dn_forward(struct sk_buff *skb) /* Ensure that we have enough space for headers */ rt = (struct dn_route *)skb_dst(skb); header_len = dn_db->use_long ? 21 : 6; - if (skb_cow(skb, LL_RESERVED_SPACE(rt->u.dst.dev)+header_len)) + if (skb_cow(skb, LL_RESERVED_SPACE(rt->dst.dev)+header_len)) goto drop; /* @@ -752,7 +752,7 @@ static int dn_forward(struct sk_buff *skb) if (++cb->hops > 30) goto drop; - skb->dev = rt->u.dst.dev; + skb->dev = rt->dst.dev; /* * If packet goes out same interface it came in on, then set @@ -792,7 +792,7 @@ static int dn_rt_bug(struct sk_buff *skb) static int dn_rt_set_next_hop(struct dn_route *rt, struct dn_fib_res *res) { struct dn_fib_info *fi = res->fi; - struct net_device *dev = rt->u.dst.dev; + struct net_device *dev = rt->dst.dev; struct neighbour *n; unsigned mss; @@ -800,25 +800,25 @@ static int dn_rt_set_next_hop(struct dn_route *rt, struct dn_fib_res *res) if (DN_FIB_RES_GW(*res) && DN_FIB_RES_NH(*res).nh_scope == RT_SCOPE_LINK) rt->rt_gateway = DN_FIB_RES_GW(*res); - memcpy(rt->u.dst.metrics, fi->fib_metrics, - sizeof(rt->u.dst.metrics)); + memcpy(rt->dst.metrics, fi->fib_metrics, + sizeof(rt->dst.metrics)); } rt->rt_type = res->type; - if (dev != NULL && rt->u.dst.neighbour == NULL) { + if (dev != NULL && rt->dst.neighbour == NULL) { n = __neigh_lookup_errno(&dn_neigh_table, &rt->rt_gateway, dev); if (IS_ERR(n)) return PTR_ERR(n); - rt->u.dst.neighbour = n; + rt->dst.neighbour = n; } - if (dst_metric(&rt->u.dst, RTAX_MTU) == 0 || - dst_metric(&rt->u.dst, RTAX_MTU) > rt->u.dst.dev->mtu) - rt->u.dst.metrics[RTAX_MTU-1] = rt->u.dst.dev->mtu; - mss = dn_mss_from_pmtu(dev, dst_mtu(&rt->u.dst)); - if (dst_metric(&rt->u.dst, RTAX_ADVMSS) == 0 || - dst_metric(&rt->u.dst, RTAX_ADVMSS) > mss) - rt->u.dst.metrics[RTAX_ADVMSS-1] = mss; + if (dst_metric(&rt->dst, RTAX_MTU) == 0 || + dst_metric(&rt->dst, RTAX_MTU) > rt->dst.dev->mtu) + rt->dst.metrics[RTAX_MTU-1] = rt->dst.dev->mtu; + mss = dn_mss_from_pmtu(dev, dst_mtu(&rt->dst)); + if (dst_metric(&rt->dst, RTAX_ADVMSS) == 0 || + dst_metric(&rt->dst, RTAX_ADVMSS) > mss) + rt->dst.metrics[RTAX_ADVMSS-1] = mss; return 0; } @@ -1096,8 +1096,8 @@ make_route: if (rt == NULL) goto e_nobufs; - atomic_set(&rt->u.dst.__refcnt, 1); - rt->u.dst.flags = DST_HOST; + atomic_set(&rt->dst.__refcnt, 1); + rt->dst.flags = DST_HOST; rt->fl.fld_src = oldflp->fld_src; rt->fl.fld_dst = oldflp->fld_dst; @@ -1113,17 +1113,17 @@ make_route: rt->rt_dst_map = fl.fld_dst; rt->rt_src_map = fl.fld_src; - rt->u.dst.dev = dev_out; + rt->dst.dev = dev_out; dev_hold(dev_out); - rt->u.dst.neighbour = neigh; + rt->dst.neighbour = neigh; neigh = NULL; - rt->u.dst.lastuse = jiffies; - rt->u.dst.output = dn_output; - rt->u.dst.input = dn_rt_bug; + rt->dst.lastuse = jiffies; + rt->dst.output = dn_output; + rt->dst.input = dn_rt_bug; rt->rt_flags = flags; if (flags & RTCF_LOCAL) - rt->u.dst.input = dn_nsp_rx; + rt->dst.input = dn_nsp_rx; err = dn_rt_set_next_hop(rt, &res); if (err) @@ -1152,7 +1152,7 @@ e_nobufs: err = -ENOBUFS; goto done; e_neighbour: - dst_free(&rt->u.dst); + dst_free(&rt->dst); goto e_nobufs; } @@ -1168,15 +1168,15 @@ static int __dn_route_output_key(struct dst_entry **pprt, const struct flowi *fl if (!(flags & MSG_TRYHARD)) { rcu_read_lock_bh(); for (rt = rcu_dereference_bh(dn_rt_hash_table[hash].chain); rt; - rt = rcu_dereference_bh(rt->u.dst.dn_next)) { + rt = rcu_dereference_bh(rt->dst.dn_next)) { if ((flp->fld_dst == rt->fl.fld_dst) && (flp->fld_src == rt->fl.fld_src) && (flp->mark == rt->fl.mark) && (rt->fl.iif == 0) && (rt->fl.oif == flp->oif)) { - dst_use(&rt->u.dst, jiffies); + dst_use(&rt->dst, jiffies); rcu_read_unlock_bh(); - *pprt = &rt->u.dst; + *pprt = &rt->dst; return 0; } } @@ -1375,29 +1375,29 @@ make_route: rt->fl.iif = in_dev->ifindex; rt->fl.mark = fl.mark; - rt->u.dst.flags = DST_HOST; - rt->u.dst.neighbour = neigh; - rt->u.dst.dev = out_dev; - rt->u.dst.lastuse = jiffies; - rt->u.dst.output = dn_rt_bug; + rt->dst.flags = DST_HOST; + rt->dst.neighbour = neigh; + rt->dst.dev = out_dev; + rt->dst.lastuse = jiffies; + rt->dst.output = dn_rt_bug; switch(res.type) { case RTN_UNICAST: - rt->u.dst.input = dn_forward; + rt->dst.input = dn_forward; break; case RTN_LOCAL: - rt->u.dst.output = dn_output; - rt->u.dst.input = dn_nsp_rx; - rt->u.dst.dev = in_dev; + rt->dst.output = dn_output; + rt->dst.input = dn_nsp_rx; + rt->dst.dev = in_dev; flags |= RTCF_LOCAL; break; default: case RTN_UNREACHABLE: case RTN_BLACKHOLE: - rt->u.dst.input = dst_discard; + rt->dst.input = dst_discard; } rt->rt_flags = flags; - if (rt->u.dst.dev) - dev_hold(rt->u.dst.dev); + if (rt->dst.dev) + dev_hold(rt->dst.dev); err = dn_rt_set_next_hop(rt, &res); if (err) @@ -1405,7 +1405,7 @@ make_route: hash = dn_hash(rt->fl.fld_src, rt->fl.fld_dst); dn_insert_route(rt, hash, &rt); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); done: if (neigh) @@ -1427,7 +1427,7 @@ e_nobufs: goto done; e_neighbour: - dst_free(&rt->u.dst); + dst_free(&rt->dst); goto done; } @@ -1442,13 +1442,13 @@ static int dn_route_input(struct sk_buff *skb) rcu_read_lock(); for(rt = rcu_dereference(dn_rt_hash_table[hash].chain); rt != NULL; - rt = rcu_dereference(rt->u.dst.dn_next)) { + rt = rcu_dereference(rt->dst.dn_next)) { if ((rt->fl.fld_src == cb->src) && (rt->fl.fld_dst == cb->dst) && (rt->fl.oif == 0) && (rt->fl.mark == skb->mark) && (rt->fl.iif == cb->iif)) { - dst_use(&rt->u.dst, jiffies); + dst_use(&rt->dst, jiffies); rcu_read_unlock(); skb_dst_set(skb, (struct dst_entry *)rt); return 0; @@ -1487,8 +1487,8 @@ static int dn_rt_fill_info(struct sk_buff *skb, u32 pid, u32 seq, r->rtm_src_len = 16; RTA_PUT(skb, RTA_SRC, 2, &rt->fl.fld_src); } - if (rt->u.dst.dev) - RTA_PUT(skb, RTA_OIF, sizeof(int), &rt->u.dst.dev->ifindex); + if (rt->dst.dev) + RTA_PUT(skb, RTA_OIF, sizeof(int), &rt->dst.dev->ifindex); /* * Note to self - change this if input routes reverse direction when * they deal only with inputs and not with replies like they do @@ -1497,11 +1497,11 @@ static int dn_rt_fill_info(struct sk_buff *skb, u32 pid, u32 seq, RTA_PUT(skb, RTA_PREFSRC, 2, &rt->rt_local_src); if (rt->rt_daddr != rt->rt_gateway) RTA_PUT(skb, RTA_GATEWAY, 2, &rt->rt_gateway); - if (rtnetlink_put_metrics(skb, rt->u.dst.metrics) < 0) + if (rtnetlink_put_metrics(skb, rt->dst.metrics) < 0) goto rtattr_failure; - expires = rt->u.dst.expires ? rt->u.dst.expires - jiffies : 0; - if (rtnl_put_cacheinfo(skb, &rt->u.dst, 0, 0, 0, expires, - rt->u.dst.error) < 0) + expires = rt->dst.expires ? rt->dst.expires - jiffies : 0; + if (rtnl_put_cacheinfo(skb, &rt->dst, 0, 0, 0, expires, + rt->dst.error) < 0) goto rtattr_failure; if (rt->fl.iif) RTA_PUT(skb, RTA_IIF, sizeof(int), &rt->fl.iif); @@ -1568,8 +1568,8 @@ static int dn_cache_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh, void local_bh_enable(); memset(cb, 0, sizeof(struct dn_skb_cb)); rt = (struct dn_route *)skb_dst(skb); - if (!err && -rt->u.dst.error) - err = rt->u.dst.error; + if (!err && -rt->dst.error) + err = rt->dst.error; } else { int oif = 0; if (rta[RTA_OIF - 1]) @@ -1583,7 +1583,7 @@ static int dn_cache_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh, void skb->dev = NULL; if (err) goto out_free; - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); if (rtm->rtm_flags & RTM_F_NOTIFY) rt->rt_flags |= RTCF_NOTIFY; @@ -1632,10 +1632,10 @@ int dn_cache_dump(struct sk_buff *skb, struct netlink_callback *cb) rcu_read_lock_bh(); for(rt = rcu_dereference_bh(dn_rt_hash_table[h].chain), idx = 0; rt; - rt = rcu_dereference_bh(rt->u.dst.dn_next), idx++) { + rt = rcu_dereference_bh(rt->dst.dn_next), idx++) { if (idx < s_idx) continue; - skb_dst_set(skb, dst_clone(&rt->u.dst)); + skb_dst_set(skb, dst_clone(&rt->dst)); if (dn_rt_fill_info(skb, NETLINK_CB(cb->skb).pid, cb->nlh->nlmsg_seq, RTM_NEWROUTE, 1, NLM_F_MULTI) <= 0) { @@ -1678,7 +1678,7 @@ static struct dn_route *dn_rt_cache_get_next(struct seq_file *seq, struct dn_rou { struct dn_rt_cache_iter_state *s = seq->private; - rt = rt->u.dst.dn_next; + rt = rt->dst.dn_next; while(!rt) { rcu_read_unlock_bh(); if (--s->bucket < 0) @@ -1719,12 +1719,12 @@ static int dn_rt_cache_seq_show(struct seq_file *seq, void *v) char buf1[DN_ASCBUF_LEN], buf2[DN_ASCBUF_LEN]; seq_printf(seq, "%-8s %-7s %-7s %04d %04d %04d\n", - rt->u.dst.dev ? rt->u.dst.dev->name : "*", + rt->dst.dev ? rt->dst.dev->name : "*", dn_addr2asc(le16_to_cpu(rt->rt_daddr), buf1), dn_addr2asc(le16_to_cpu(rt->rt_saddr), buf2), - atomic_read(&rt->u.dst.__refcnt), - rt->u.dst.__use, - (int) dst_metric(&rt->u.dst, RTAX_RTT)); + atomic_read(&rt->dst.__refcnt), + rt->dst.__use, + (int) dst_metric(&rt->dst, RTAX_RTT)); return 0; } diff --git a/net/ethernet/eth.c b/net/ethernet/eth.c index 61ec032..215c839 100644 --- a/net/ethernet/eth.c +++ b/net/ethernet/eth.c @@ -158,7 +158,6 @@ EXPORT_SYMBOL(eth_rebuild_header); __be16 eth_type_trans(struct sk_buff *skb, struct net_device *dev) { struct ethhdr *eth; - unsigned char *rawp; skb->dev = dev; skb_reset_mac_header(skb); @@ -199,15 +198,13 @@ __be16 eth_type_trans(struct sk_buff *skb, struct net_device *dev) if (ntohs(eth->h_proto) >= 1536) return eth->h_proto; - rawp = skb->data; - /* * This is a magic hack to spot IPX packets. Older Novell breaks * the protocol design and runs IPX over 802.3 without an 802.2 LLC * layer. We look for FFFF which isn't a used 802.2 SSAP/DSAP. This * won't work for fault tolerant netware but does for the rest. */ - if (*(unsigned short *)rawp == 0xFFFF) + if (skb->len >= 2 && *(unsigned short *)(skb->data) == 0xFFFF) return htons(ETH_P_802_3); /* diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 551ce56..d99e7e0 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -1100,7 +1100,7 @@ static int inet_sk_reselect_saddr(struct sock *sk) if (err) return err; - sk_setup_caps(sk, &rt->u.dst); + sk_setup_caps(sk, &rt->dst); new_saddr = rt->rt_src; @@ -1166,7 +1166,7 @@ int inet_sk_rebuild_header(struct sock *sk) err = ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 0); } if (!err) - sk_setup_caps(sk, &rt->u.dst); + sk_setup_caps(sk, &rt->dst); else { /* Routing failed... */ sk->sk_route_caps = 0; diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index 917d2d6..cf78f41 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -427,7 +427,7 @@ static int arp_filter(__be32 sip, __be32 tip, struct net_device *dev) if (ip_route_output_key(net, &rt, &fl) < 0) return 1; - if (rt->u.dst.dev != dev) { + if (rt->dst.dev != dev) { NET_INC_STATS_BH(net, LINUX_MIB_ARPFILTER); flag = 1; } @@ -532,7 +532,7 @@ static inline int arp_fwd_proxy(struct in_device *in_dev, struct in_device *out_dev; int imi, omi = -1; - if (rt->u.dst.dev == dev) + if (rt->dst.dev == dev) return 0; if (!IN_DEV_PROXY_ARP(in_dev)) @@ -545,7 +545,7 @@ static inline int arp_fwd_proxy(struct in_device *in_dev, /* place to check for proxy_arp for routes */ - out_dev = __in_dev_get_rcu(rt->u.dst.dev); + out_dev = __in_dev_get_rcu(rt->dst.dev); if (out_dev) omi = IN_DEV_MEDIUM_ID(out_dev); @@ -576,7 +576,7 @@ static inline int arp_fwd_pvlan(struct in_device *in_dev, __be32 sip, __be32 tip) { /* Private VLAN is only concerned about the same ethernet segment */ - if (rt->u.dst.dev != dev) + if (rt->dst.dev != dev) return 0; /* Don't reply on self probes (often done by windowz boxes)*/ @@ -1042,7 +1042,7 @@ static int arp_req_set(struct net *net, struct arpreq *r, struct rtable * rt; if ((err = ip_route_output_key(net, &rt, &fl)) != 0) return err; - dev = rt->u.dst.dev; + dev = rt->dst.dev; ip_rt_put(rt); if (!dev) return -EINVAL; @@ -1149,7 +1149,7 @@ static int arp_req_delete(struct net *net, struct arpreq *r, struct rtable * rt; if ((err = ip_route_output_key(net, &rt, &fl)) != 0) return err; - dev = rt->u.dst.dev; + dev = rt->dst.dev; ip_rt_put(rt); if (!dev) return -EINVAL; diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c index fb24658..fe3daa7 100644 --- a/net/ipv4/datagram.c +++ b/net/ipv4/datagram.c @@ -69,7 +69,7 @@ int ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) sk->sk_state = TCP_ESTABLISHED; inet->inet_id = jiffies; - sk_dst_set(sk, &rt->u.dst); + sk_dst_set(sk, &rt->dst); return(0); } diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index bdb6c71..7569b21 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -271,7 +271,7 @@ int xrlim_allow(struct dst_entry *dst, int timeout) static inline int icmpv4_xrlim_allow(struct net *net, struct rtable *rt, int type, int code) { - struct dst_entry *dst = &rt->u.dst; + struct dst_entry *dst = &rt->dst; int rc = 1; if (type > NR_ICMP_TYPES) @@ -327,7 +327,7 @@ static void icmp_push_reply(struct icmp_bxm *icmp_param, struct sock *sk; struct sk_buff *skb; - sk = icmp_sk(dev_net((*rt)->u.dst.dev)); + sk = icmp_sk(dev_net((*rt)->dst.dev)); if (ip_append_data(sk, icmp_glue_bits, icmp_param, icmp_param->data_len+icmp_param->head_len, icmp_param->head_len, @@ -359,7 +359,7 @@ static void icmp_reply(struct icmp_bxm *icmp_param, struct sk_buff *skb) { struct ipcm_cookie ipc; struct rtable *rt = skb_rtable(skb); - struct net *net = dev_net(rt->u.dst.dev); + struct net *net = dev_net(rt->dst.dev); struct sock *sk; struct inet_sock *inet; __be32 daddr; @@ -427,7 +427,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info) if (!rt) goto out; - net = dev_net(rt->u.dst.dev); + net = dev_net(rt->dst.dev); /* * Find the original header. It is expected to be valid, of course. @@ -596,9 +596,9 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info) /* Ugh! */ orefdst = skb_in->_skb_refdst; /* save old refdst */ err = ip_route_input(skb_in, fl.fl4_dst, fl.fl4_src, - RT_TOS(tos), rt2->u.dst.dev); + RT_TOS(tos), rt2->dst.dev); - dst_release(&rt2->u.dst); + dst_release(&rt2->dst); rt2 = skb_rtable(skb_in); skb_in->_skb_refdst = orefdst; /* restore old refdst */ } @@ -610,7 +610,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info) XFRM_LOOKUP_ICMP); switch (err) { case 0: - dst_release(&rt->u.dst); + dst_release(&rt->dst); rt = rt2; break; case -EPERM: @@ -629,7 +629,7 @@ route_done: /* RFC says return as much as we can without exceeding 576 bytes. */ - room = dst_mtu(&rt->u.dst); + room = dst_mtu(&rt->dst); if (room > 576) room = 576; room -= sizeof(struct iphdr) + icmp_param.replyopts.optlen; @@ -972,7 +972,7 @@ int icmp_rcv(struct sk_buff *skb) { struct icmphdr *icmph; struct rtable *rt = skb_rtable(skb); - struct net *net = dev_net(rt->u.dst.dev); + struct net *net = dev_net(rt->dst.dev); if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) { struct sec_path *sp = skb_sec_path(skb); diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index 3294f54..b5580d4 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -312,7 +312,7 @@ static struct sk_buff *igmpv3_newpack(struct net_device *dev, int size) return NULL; } - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); skb->dev = dev; skb_reserve(skb, LL_RESERVED_SPACE(dev)); @@ -330,7 +330,7 @@ static struct sk_buff *igmpv3_newpack(struct net_device *dev, int size) pip->saddr = rt->rt_src; pip->protocol = IPPROTO_IGMP; pip->tot_len = 0; /* filled in later */ - ip_select_ident(pip, &rt->u.dst, NULL); + ip_select_ident(pip, &rt->dst, NULL); ((u8*)&pip[1])[0] = IPOPT_RA; ((u8*)&pip[1])[1] = 4; ((u8*)&pip[1])[2] = 0; @@ -660,7 +660,7 @@ static int igmp_send_report(struct in_device *in_dev, struct ip_mc_list *pmc, return -1; } - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); skb_reserve(skb, LL_RESERVED_SPACE(dev)); @@ -676,7 +676,7 @@ static int igmp_send_report(struct in_device *in_dev, struct ip_mc_list *pmc, iph->daddr = dst; iph->saddr = rt->rt_src; iph->protocol = IPPROTO_IGMP; - ip_select_ident(iph, &rt->u.dst, NULL); + ip_select_ident(iph, &rt->dst, NULL); ((u8*)&iph[1])[0] = IPOPT_RA; ((u8*)&iph[1])[1] = 4; ((u8*)&iph[1])[2] = 0; @@ -1425,7 +1425,7 @@ static struct in_device *ip_mc_find_dev(struct net *net, struct ip_mreqn *imr) } if (!dev && !ip_route_output_key(net, &rt, &fl)) { - dev = rt->u.dst.dev; + dev = rt->dst.dev; ip_rt_put(rt); } if (dev) { diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 70eb350..57c9e4d 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -383,7 +383,7 @@ struct dst_entry *inet_csk_route_req(struct sock *sk, goto no_route; if (opt && opt->is_strictroute && rt->rt_dst != rt->rt_gateway) goto route_err; - return &rt->u.dst; + return &rt->dst; route_err: ip_rt_put(rt); diff --git a/net/ipv4/ip_forward.c b/net/ipv4/ip_forward.c index 56cdf68..99461f0 100644 --- a/net/ipv4/ip_forward.c +++ b/net/ipv4/ip_forward.c @@ -87,16 +87,16 @@ int ip_forward(struct sk_buff *skb) if (opt->is_strictroute && rt->rt_dst != rt->rt_gateway) goto sr_failed; - if (unlikely(skb->len > dst_mtu(&rt->u.dst) && !skb_is_gso(skb) && + if (unlikely(skb->len > dst_mtu(&rt->dst) && !skb_is_gso(skb) && (ip_hdr(skb)->frag_off & htons(IP_DF))) && !skb->local_df) { - IP_INC_STATS(dev_net(rt->u.dst.dev), IPSTATS_MIB_FRAGFAILS); + IP_INC_STATS(dev_net(rt->dst.dev), IPSTATS_MIB_FRAGFAILS); icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, - htonl(dst_mtu(&rt->u.dst))); + htonl(dst_mtu(&rt->dst))); goto drop; } /* We are about to mangle packet. Copy it! */ - if (skb_cow(skb, LL_RESERVED_SPACE(rt->u.dst.dev)+rt->u.dst.header_len)) + if (skb_cow(skb, LL_RESERVED_SPACE(rt->dst.dev)+rt->dst.header_len)) goto drop; iph = ip_hdr(skb); @@ -113,7 +113,7 @@ int ip_forward(struct sk_buff *skb) skb->priority = rt_tos2priority(iph->tos); return NF_HOOK(NFPROTO_IPV4, NF_INET_FORWARD, skb, skb->dev, - rt->u.dst.dev, ip_forward_finish); + rt->dst.dev, ip_forward_finish); sr_failed: /* diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c index 32618e1..749e548 100644 --- a/net/ipv4/ip_gre.c +++ b/net/ipv4/ip_gre.c @@ -745,7 +745,7 @@ static netdev_tx_t ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev goto tx_error; } } - tdev = rt->u.dst.dev; + tdev = rt->dst.dev; if (tdev == dev) { ip_rt_put(rt); @@ -755,7 +755,7 @@ static netdev_tx_t ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev df = tiph->frag_off; if (df) - mtu = dst_mtu(&rt->u.dst) - dev->hard_header_len - tunnel->hlen; + mtu = dst_mtu(&rt->dst) - dev->hard_header_len - tunnel->hlen; else mtu = skb_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu; @@ -803,7 +803,7 @@ static netdev_tx_t ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev tunnel->err_count = 0; } - max_headroom = LL_RESERVED_SPACE(tdev) + gre_hlen + rt->u.dst.header_len; + max_headroom = LL_RESERVED_SPACE(tdev) + gre_hlen + rt->dst.header_len; if (skb_headroom(skb) < max_headroom || skb_shared(skb)|| (skb_cloned(skb) && !skb_clone_writable(skb, 0))) { @@ -830,7 +830,7 @@ static netdev_tx_t ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED | IPSKB_REROUTED); skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* * Push down and install the IPIP header. @@ -853,7 +853,7 @@ static netdev_tx_t ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev iph->ttl = ((struct ipv6hdr *)old_iph)->hop_limit; #endif else - iph->ttl = dst_metric(&rt->u.dst, RTAX_HOPLIMIT); + iph->ttl = dst_metric(&rt->dst, RTAX_HOPLIMIT); } ((__be16 *)(iph + 1))[0] = tunnel->parms.o_flags; @@ -915,7 +915,7 @@ static int ipgre_tunnel_bind_dev(struct net_device *dev) .proto = IPPROTO_GRE }; struct rtable *rt; if (!ip_route_output_key(dev_net(dev), &rt, &fl)) { - tdev = rt->u.dst.dev; + tdev = rt->dst.dev; ip_rt_put(rt); } @@ -1174,7 +1174,7 @@ static int ipgre_open(struct net_device *dev) struct rtable *rt; if (ip_route_output_key(dev_net(dev), &rt, &fl)) return -EADDRNOTAVAIL; - dev = rt->u.dst.dev; + dev = rt->dst.dev; ip_rt_put(rt); if (__in_dev_get_rtnl(dev) == NULL) return -EADDRNOTAVAIL; diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c index 08a3b12..db47a5a 100644 --- a/net/ipv4/ip_input.c +++ b/net/ipv4/ip_input.c @@ -356,10 +356,10 @@ static int ip_rcv_finish(struct sk_buff *skb) rt = skb_rtable(skb); if (rt->rt_type == RTN_MULTICAST) { - IP_UPD_PO_STATS_BH(dev_net(rt->u.dst.dev), IPSTATS_MIB_INMCAST, + IP_UPD_PO_STATS_BH(dev_net(rt->dst.dev), IPSTATS_MIB_INMCAST, skb->len); } else if (rt->rt_type == RTN_BROADCAST) - IP_UPD_PO_STATS_BH(dev_net(rt->u.dst.dev), IPSTATS_MIB_INBCAST, + IP_UPD_PO_STATS_BH(dev_net(rt->dst.dev), IPSTATS_MIB_INBCAST, skb->len); return dst_input(skb); diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 9a4a6c9..6cbeb2e 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -151,15 +151,15 @@ int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk, iph->version = 4; iph->ihl = 5; iph->tos = inet->tos; - if (ip_dont_fragment(sk, &rt->u.dst)) + if (ip_dont_fragment(sk, &rt->dst)) iph->frag_off = htons(IP_DF); else iph->frag_off = 0; - iph->ttl = ip_select_ttl(inet, &rt->u.dst); + iph->ttl = ip_select_ttl(inet, &rt->dst); iph->daddr = rt->rt_dst; iph->saddr = rt->rt_src; iph->protocol = sk->sk_protocol; - ip_select_ident(iph, &rt->u.dst, sk); + ip_select_ident(iph, &rt->dst, sk); if (opt && opt->optlen) { iph->ihl += opt->optlen>>2; @@ -240,7 +240,7 @@ int ip_mc_output(struct sk_buff *skb) { struct sock *sk = skb->sk; struct rtable *rt = skb_rtable(skb); - struct net_device *dev = rt->u.dst.dev; + struct net_device *dev = rt->dst.dev; /* * If the indicated interface is up and running, send the packet. @@ -359,9 +359,9 @@ int ip_queue_xmit(struct sk_buff *skb) if (ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 0)) goto no_route; } - sk_setup_caps(sk, &rt->u.dst); + sk_setup_caps(sk, &rt->dst); } - skb_dst_set_noref(skb, &rt->u.dst); + skb_dst_set_noref(skb, &rt->dst); packet_routed: if (opt && opt->is_strictroute && rt->rt_dst != rt->rt_gateway) @@ -372,11 +372,11 @@ packet_routed: skb_reset_network_header(skb); iph = ip_hdr(skb); *((__be16 *)iph) = htons((4 << 12) | (5 << 8) | (inet->tos & 0xff)); - if (ip_dont_fragment(sk, &rt->u.dst) && !skb->local_df) + if (ip_dont_fragment(sk, &rt->dst) && !skb->local_df) iph->frag_off = htons(IP_DF); else iph->frag_off = 0; - iph->ttl = ip_select_ttl(inet, &rt->u.dst); + iph->ttl = ip_select_ttl(inet, &rt->dst); iph->protocol = sk->sk_protocol; iph->saddr = rt->rt_src; iph->daddr = rt->rt_dst; @@ -387,7 +387,7 @@ packet_routed: ip_options_build(skb, opt, inet->inet_daddr, rt, 0); } - ip_select_ident_more(iph, &rt->u.dst, sk, + ip_select_ident_more(iph, &rt->dst, sk, (skb_shinfo(skb)->gso_segs ?: 1) - 1); skb->priority = sk->sk_priority; @@ -452,7 +452,7 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) struct rtable *rt = skb_rtable(skb); int err = 0; - dev = rt->u.dst.dev; + dev = rt->dst.dev; /* * Point into the IP datagram header. @@ -473,7 +473,7 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) */ hlen = iph->ihl * 4; - mtu = dst_mtu(&rt->u.dst) - hlen; /* Size of data space */ + mtu = dst_mtu(&rt->dst) - hlen; /* Size of data space */ #ifdef CONFIG_BRIDGE_NETFILTER if (skb->nf_bridge) mtu -= nf_bridge_mtu_reduction(skb); @@ -586,7 +586,7 @@ slow_path: * we need to make room for the encapsulating header */ pad = nf_bridge_pad(skb); - ll_rs = LL_RESERVED_SPACE_EXTRA(rt->u.dst.dev, pad); + ll_rs = LL_RESERVED_SPACE_EXTRA(rt->dst.dev, pad); mtu -= pad; /* @@ -833,13 +833,13 @@ int ip_append_data(struct sock *sk, */ *rtp = NULL; inet->cork.fragsize = mtu = inet->pmtudisc == IP_PMTUDISC_PROBE ? - rt->u.dst.dev->mtu : - dst_mtu(rt->u.dst.path); - inet->cork.dst = &rt->u.dst; + rt->dst.dev->mtu : + dst_mtu(rt->dst.path); + inet->cork.dst = &rt->dst; inet->cork.length = 0; sk->sk_sndmsg_page = NULL; sk->sk_sndmsg_off = 0; - if ((exthdrlen = rt->u.dst.header_len) != 0) { + if ((exthdrlen = rt->dst.header_len) != 0) { length += exthdrlen; transhdrlen += exthdrlen; } @@ -852,7 +852,7 @@ int ip_append_data(struct sock *sk, exthdrlen = 0; mtu = inet->cork.fragsize; } - hh_len = LL_RESERVED_SPACE(rt->u.dst.dev); + hh_len = LL_RESERVED_SPACE(rt->dst.dev); fragheaderlen = sizeof(struct iphdr) + (opt ? opt->optlen : 0); maxfraglen = ((mtu - fragheaderlen) & ~7) + fragheaderlen; @@ -869,14 +869,14 @@ int ip_append_data(struct sock *sk, */ if (transhdrlen && length + fragheaderlen <= mtu && - rt->u.dst.dev->features & NETIF_F_V4_CSUM && + rt->dst.dev->features & NETIF_F_V4_CSUM && !exthdrlen) csummode = CHECKSUM_PARTIAL; inet->cork.length += length; if (((length> mtu) || !skb_queue_empty(&sk->sk_write_queue)) && (sk->sk_protocol == IPPROTO_UDP) && - (rt->u.dst.dev->features & NETIF_F_UFO)) { + (rt->dst.dev->features & NETIF_F_UFO)) { err = ip_ufo_append_data(sk, getfrag, from, length, hh_len, fragheaderlen, transhdrlen, mtu, flags); @@ -924,7 +924,7 @@ alloc_new_skb: fraglen = datalen + fragheaderlen; if ((flags & MSG_MORE) && - !(rt->u.dst.dev->features&NETIF_F_SG)) + !(rt->dst.dev->features&NETIF_F_SG)) alloclen = mtu; else alloclen = datalen + fragheaderlen; @@ -935,7 +935,7 @@ alloc_new_skb: * the last. */ if (datalen == length + fraggap) - alloclen += rt->u.dst.trailer_len; + alloclen += rt->dst.trailer_len; if (transhdrlen) { skb = sock_alloc_send_skb(sk, @@ -1008,7 +1008,7 @@ alloc_new_skb: if (copy > length) copy = length; - if (!(rt->u.dst.dev->features&NETIF_F_SG)) { + if (!(rt->dst.dev->features&NETIF_F_SG)) { unsigned int off; off = skb->len; @@ -1103,10 +1103,10 @@ ssize_t ip_append_page(struct sock *sk, struct page *page, if (inet->cork.flags & IPCORK_OPT) opt = inet->cork.opt; - if (!(rt->u.dst.dev->features&NETIF_F_SG)) + if (!(rt->dst.dev->features&NETIF_F_SG)) return -EOPNOTSUPP; - hh_len = LL_RESERVED_SPACE(rt->u.dst.dev); + hh_len = LL_RESERVED_SPACE(rt->dst.dev); mtu = inet->cork.fragsize; fragheaderlen = sizeof(struct iphdr) + (opt ? opt->optlen : 0); @@ -1122,7 +1122,7 @@ ssize_t ip_append_page(struct sock *sk, struct page *page, inet->cork.length += size; if ((sk->sk_protocol == IPPROTO_UDP) && - (rt->u.dst.dev->features & NETIF_F_UFO)) { + (rt->dst.dev->features & NETIF_F_UFO)) { skb_shinfo(skb)->gso_size = mtu - fragheaderlen; skb_shinfo(skb)->gso_type = SKB_GSO_UDP; } @@ -1274,8 +1274,8 @@ int ip_push_pending_frames(struct sock *sk) * If local_df is set too, we still allow to fragment this frame * locally. */ if (inet->pmtudisc >= IP_PMTUDISC_DO || - (skb->len <= dst_mtu(&rt->u.dst) && - ip_dont_fragment(sk, &rt->u.dst))) + (skb->len <= dst_mtu(&rt->dst) && + ip_dont_fragment(sk, &rt->dst))) df = htons(IP_DF); if (inet->cork.flags & IPCORK_OPT) @@ -1284,7 +1284,7 @@ int ip_push_pending_frames(struct sock *sk) if (rt->rt_type == RTN_MULTICAST) ttl = inet->mc_ttl; else - ttl = ip_select_ttl(inet, &rt->u.dst); + ttl = ip_select_ttl(inet, &rt->dst); iph = (struct iphdr *)skb->data; iph->version = 4; @@ -1295,7 +1295,7 @@ int ip_push_pending_frames(struct sock *sk) } iph->tos = inet->tos; iph->frag_off = df; - ip_select_ident(iph, &rt->u.dst, sk); + ip_select_ident(iph, &rt->dst, sk); iph->ttl = ttl; iph->protocol = sk->sk_protocol; iph->saddr = rt->rt_src; @@ -1308,7 +1308,7 @@ int ip_push_pending_frames(struct sock *sk) * on dst refcount */ inet->cork.dst = NULL; - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); if (iph->protocol == IPPROTO_ICMP) icmp_out_count(net, ((struct icmphdr *) diff --git a/net/ipv4/ipip.c b/net/ipv4/ipip.c index 7fd6367..ec03673 100644 --- a/net/ipv4/ipip.c +++ b/net/ipv4/ipip.c @@ -435,7 +435,7 @@ static netdev_tx_t ipip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev) goto tx_error_icmp; } } - tdev = rt->u.dst.dev; + tdev = rt->dst.dev; if (tdev == dev) { ip_rt_put(rt); @@ -446,7 +446,7 @@ static netdev_tx_t ipip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev) df |= old_iph->frag_off & htons(IP_DF); if (df) { - mtu = dst_mtu(&rt->u.dst) - sizeof(struct iphdr); + mtu = dst_mtu(&rt->dst) - sizeof(struct iphdr); if (mtu < 68) { stats->collisions++; @@ -503,7 +503,7 @@ static netdev_tx_t ipip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev) IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED | IPSKB_REROUTED); skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* * Push down and install the IPIP header. @@ -552,7 +552,7 @@ static void ipip_tunnel_bind_dev(struct net_device *dev) .proto = IPPROTO_IPIP }; struct rtable *rt; if (!ip_route_output_key(dev_net(dev), &rt, &fl)) { - tdev = rt->u.dst.dev; + tdev = rt->dst.dev; ip_rt_put(rt); } dev->flags |= IFF_POINTOPOINT; diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c index 856123f..8418afc 100644 --- a/net/ipv4/ipmr.c +++ b/net/ipv4/ipmr.c @@ -1551,9 +1551,9 @@ static void ipmr_queue_xmit(struct net *net, struct mr_table *mrt, goto out_free; } - dev = rt->u.dst.dev; + dev = rt->dst.dev; - if (skb->len+encap > dst_mtu(&rt->u.dst) && (ntohs(iph->frag_off) & IP_DF)) { + if (skb->len+encap > dst_mtu(&rt->dst) && (ntohs(iph->frag_off) & IP_DF)) { /* Do not fragment multicasts. Alas, IPv4 does not allow to send ICMP, so that packets will disappear to blackhole. @@ -1564,7 +1564,7 @@ static void ipmr_queue_xmit(struct net *net, struct mr_table *mrt, goto out_free; } - encap += LL_RESERVED_SPACE(dev) + rt->u.dst.header_len; + encap += LL_RESERVED_SPACE(dev) + rt->dst.header_len; if (skb_cow(skb, encap)) { ip_rt_put(rt); @@ -1575,7 +1575,7 @@ static void ipmr_queue_xmit(struct net *net, struct mr_table *mrt, vif->bytes_out += skb->len; skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); ip_decrease_ttl(ip_hdr(skb)); /* FIXME: forward and output firewalls used to be called here. diff --git a/net/ipv4/netfilter.c b/net/ipv4/netfilter.c index 07de855..cfbc79a 100644 --- a/net/ipv4/netfilter.c +++ b/net/ipv4/netfilter.c @@ -43,7 +43,7 @@ int ip_route_me_harder(struct sk_buff *skb, unsigned addr_type) /* Drop old route. */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); } else { /* non-local src, find valid iif to satisfy * rp-filter when calling ip_route_input. */ @@ -53,11 +53,11 @@ int ip_route_me_harder(struct sk_buff *skb, unsigned addr_type) orefdst = skb->_skb_refdst; if (ip_route_input(skb, iph->daddr, iph->saddr, - RT_TOS(iph->tos), rt->u.dst.dev) != 0) { - dst_release(&rt->u.dst); + RT_TOS(iph->tos), rt->dst.dev) != 0) { + dst_release(&rt->dst); return -1; } - dst_release(&rt->u.dst); + dst_release(&rt->dst); refdst_drop(orefdst); } diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 66cc3be..009a7b2 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -325,24 +325,24 @@ static int raw_send_hdrinc(struct sock *sk, void *from, size_t length, int err; struct rtable *rt = *rtp; - if (length > rt->u.dst.dev->mtu) { + if (length > rt->dst.dev->mtu) { ip_local_error(sk, EMSGSIZE, rt->rt_dst, inet->inet_dport, - rt->u.dst.dev->mtu); + rt->dst.dev->mtu); return -EMSGSIZE; } if (flags&MSG_PROBE) goto out; skb = sock_alloc_send_skb(sk, - length + LL_ALLOCATED_SPACE(rt->u.dst.dev) + 15, + length + LL_ALLOCATED_SPACE(rt->dst.dev) + 15, flags & MSG_DONTWAIT, &err); if (skb == NULL) goto error; - skb_reserve(skb, LL_RESERVED_SPACE(rt->u.dst.dev)); + skb_reserve(skb, LL_RESERVED_SPACE(rt->dst.dev)); skb->priority = sk->sk_priority; skb->mark = sk->sk_mark; - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); *rtp = NULL; skb_reset_network_header(skb); @@ -375,7 +375,7 @@ static int raw_send_hdrinc(struct sock *sk, void *from, size_t length, iph->check = 0; iph->tot_len = htons(length); if (!iph->id) - ip_select_ident(iph, &rt->u.dst, NULL); + ip_select_ident(iph, &rt->dst, NULL); iph->check = ip_fast_csum((unsigned char *)iph, iph->ihl); } @@ -384,7 +384,7 @@ static int raw_send_hdrinc(struct sock *sk, void *from, size_t length, skb_transport_header(skb))->type); err = NF_HOOK(NFPROTO_IPV4, NF_INET_LOCAL_OUT, skb, NULL, - rt->u.dst.dev, dst_output); + rt->dst.dev, dst_output); if (err > 0) err = net_xmit_errno(err); if (err) @@ -606,7 +606,7 @@ out: return len; do_confirm: - dst_confirm(&rt->u.dst); + dst_confirm(&rt->dst); if (!(msg->msg_flags & MSG_PROBE) || len) goto back_from_confirm; err = 0; diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 883b5c7..a291edb 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -286,10 +286,10 @@ static struct rtable *rt_cache_get_first(struct seq_file *seq) rcu_read_lock_bh(); r = rcu_dereference_bh(rt_hash_table[st->bucket].chain); while (r) { - if (dev_net(r->u.dst.dev) == seq_file_net(seq) && + if (dev_net(r->dst.dev) == seq_file_net(seq) && r->rt_genid == st->genid) return r; - r = rcu_dereference_bh(r->u.dst.rt_next); + r = rcu_dereference_bh(r->dst.rt_next); } rcu_read_unlock_bh(); } @@ -301,7 +301,7 @@ static struct rtable *__rt_cache_get_next(struct seq_file *seq, { struct rt_cache_iter_state *st = seq->private; - r = r->u.dst.rt_next; + r = r->dst.rt_next; while (!r) { rcu_read_unlock_bh(); do { @@ -319,7 +319,7 @@ static struct rtable *rt_cache_get_next(struct seq_file *seq, { struct rt_cache_iter_state *st = seq->private; while ((r = __rt_cache_get_next(seq, r)) != NULL) { - if (dev_net(r->u.dst.dev) != seq_file_net(seq)) + if (dev_net(r->dst.dev) != seq_file_net(seq)) continue; if (r->rt_genid == st->genid) break; @@ -377,19 +377,19 @@ static int rt_cache_seq_show(struct seq_file *seq, void *v) seq_printf(seq, "%s\t%08X\t%08X\t%8X\t%d\t%u\t%d\t" "%08X\t%d\t%u\t%u\t%02X\t%d\t%1d\t%08X%n", - r->u.dst.dev ? r->u.dst.dev->name : "*", + r->dst.dev ? r->dst.dev->name : "*", (__force u32)r->rt_dst, (__force u32)r->rt_gateway, - r->rt_flags, atomic_read(&r->u.dst.__refcnt), - r->u.dst.__use, 0, (__force u32)r->rt_src, - (dst_metric(&r->u.dst, RTAX_ADVMSS) ? - (int)dst_metric(&r->u.dst, RTAX_ADVMSS) + 40 : 0), - dst_metric(&r->u.dst, RTAX_WINDOW), - (int)((dst_metric(&r->u.dst, RTAX_RTT) >> 3) + - dst_metric(&r->u.dst, RTAX_RTTVAR)), + r->rt_flags, atomic_read(&r->dst.__refcnt), + r->dst.__use, 0, (__force u32)r->rt_src, + (dst_metric(&r->dst, RTAX_ADVMSS) ? + (int)dst_metric(&r->dst, RTAX_ADVMSS) + 40 : 0), + dst_metric(&r->dst, RTAX_WINDOW), + (int)((dst_metric(&r->dst, RTAX_RTT) >> 3) + + dst_metric(&r->dst, RTAX_RTTVAR)), r->fl.fl4_tos, - r->u.dst.hh ? atomic_read(&r->u.dst.hh->hh_refcnt) : -1, - r->u.dst.hh ? (r->u.dst.hh->hh_output == + r->dst.hh ? atomic_read(&r->dst.hh->hh_refcnt) : -1, + r->dst.hh ? (r->dst.hh->hh_output == dev_queue_xmit) : 0, r->rt_spec_dst, &len); @@ -608,13 +608,13 @@ static inline int ip_rt_proc_init(void) static inline void rt_free(struct rtable *rt) { - call_rcu_bh(&rt->u.dst.rcu_head, dst_rcu_free); + call_rcu_bh(&rt->dst.rcu_head, dst_rcu_free); } static inline void rt_drop(struct rtable *rt) { ip_rt_put(rt); - call_rcu_bh(&rt->u.dst.rcu_head, dst_rcu_free); + call_rcu_bh(&rt->dst.rcu_head, dst_rcu_free); } static inline int rt_fast_clean(struct rtable *rth) @@ -622,13 +622,13 @@ static inline int rt_fast_clean(struct rtable *rth) /* Kill broadcast/multicast entries very aggresively, if they collide in hash table with more useful entries */ return (rth->rt_flags & (RTCF_BROADCAST | RTCF_MULTICAST)) && - rth->fl.iif && rth->u.dst.rt_next; + rth->fl.iif && rth->dst.rt_next; } static inline int rt_valuable(struct rtable *rth) { return (rth->rt_flags & (RTCF_REDIRECTED | RTCF_NOTIFY)) || - rth->u.dst.expires; + rth->dst.expires; } static int rt_may_expire(struct rtable *rth, unsigned long tmo1, unsigned long tmo2) @@ -636,15 +636,15 @@ static int rt_may_expire(struct rtable *rth, unsigned long tmo1, unsigned long t unsigned long age; int ret = 0; - if (atomic_read(&rth->u.dst.__refcnt)) + if (atomic_read(&rth->dst.__refcnt)) goto out; ret = 1; - if (rth->u.dst.expires && - time_after_eq(jiffies, rth->u.dst.expires)) + if (rth->dst.expires && + time_after_eq(jiffies, rth->dst.expires)) goto out; - age = jiffies - rth->u.dst.lastuse; + age = jiffies - rth->dst.lastuse; ret = 0; if ((age <= tmo1 && !rt_fast_clean(rth)) || (age <= tmo2 && rt_valuable(rth))) @@ -660,7 +660,7 @@ out: return ret; */ static inline u32 rt_score(struct rtable *rt) { - u32 score = jiffies - rt->u.dst.lastuse; + u32 score = jiffies - rt->dst.lastuse; score = ~score & ~(3<<30); @@ -700,12 +700,12 @@ static inline int compare_keys(struct flowi *fl1, struct flowi *fl2) static inline int compare_netns(struct rtable *rt1, struct rtable *rt2) { - return net_eq(dev_net(rt1->u.dst.dev), dev_net(rt2->u.dst.dev)); + return net_eq(dev_net(rt1->dst.dev), dev_net(rt2->dst.dev)); } static inline int rt_is_expired(struct rtable *rth) { - return rth->rt_genid != rt_genid(dev_net(rth->u.dst.dev)); + return rth->rt_genid != rt_genid(dev_net(rth->dst.dev)); } /* @@ -734,7 +734,7 @@ static void rt_do_flush(int process_context) rth = rt_hash_table[i].chain; /* defer releasing the head of the list after spin_unlock */ - for (tail = rth; tail; tail = tail->u.dst.rt_next) + for (tail = rth; tail; tail = tail->dst.rt_next) if (!rt_is_expired(tail)) break; if (rth != tail) @@ -743,9 +743,9 @@ static void rt_do_flush(int process_context) /* call rt_free on entries after the tail requiring flush */ prev = &rt_hash_table[i].chain; for (p = *prev; p; p = next) { - next = p->u.dst.rt_next; + next = p->dst.rt_next; if (!rt_is_expired(p)) { - prev = &p->u.dst.rt_next; + prev = &p->dst.rt_next; } else { *prev = next; rt_free(p); @@ -760,7 +760,7 @@ static void rt_do_flush(int process_context) spin_unlock_bh(rt_hash_lock_addr(i)); for (; rth != tail; rth = next) { - next = rth->u.dst.rt_next; + next = rth->dst.rt_next; rt_free(rth); } } @@ -791,7 +791,7 @@ static int has_noalias(const struct rtable *head, const struct rtable *rth) while (aux != rth) { if (compare_hash_inputs(&aux->fl, &rth->fl)) return 0; - aux = aux->u.dst.rt_next; + aux = aux->dst.rt_next; } return ONE; } @@ -831,18 +831,18 @@ static void rt_check_expire(void) length = 0; spin_lock_bh(rt_hash_lock_addr(i)); while ((rth = *rthp) != NULL) { - prefetch(rth->u.dst.rt_next); + prefetch(rth->dst.rt_next); if (rt_is_expired(rth)) { - *rthp = rth->u.dst.rt_next; + *rthp = rth->dst.rt_next; rt_free(rth); continue; } - if (rth->u.dst.expires) { + if (rth->dst.expires) { /* Entry is expired even if it is in use */ - if (time_before_eq(jiffies, rth->u.dst.expires)) { + if (time_before_eq(jiffies, rth->dst.expires)) { nofree: tmo >>= 1; - rthp = &rth->u.dst.rt_next; + rthp = &rth->dst.rt_next; /* * We only count entries on * a chain with equal hash inputs once @@ -858,7 +858,7 @@ nofree: goto nofree; /* Cleanup aged off entries. */ - *rthp = rth->u.dst.rt_next; + *rthp = rth->dst.rt_next; rt_free(rth); } spin_unlock_bh(rt_hash_lock_addr(i)); @@ -999,10 +999,10 @@ static int rt_garbage_collect(struct dst_ops *ops) if (!rt_is_expired(rth) && !rt_may_expire(rth, tmo, expire)) { tmo >>= 1; - rthp = &rth->u.dst.rt_next; + rthp = &rth->dst.rt_next; continue; } - *rthp = rth->u.dst.rt_next; + *rthp = rth->dst.rt_next; rt_free(rth); goal--; } @@ -1068,7 +1068,7 @@ static int slow_chain_length(const struct rtable *head) while (rth) { length += has_noalias(head, rth); - rth = rth->u.dst.rt_next; + rth = rth->dst.rt_next; } return length >> FRACT_BITS; } @@ -1090,7 +1090,7 @@ restart: candp = NULL; now = jiffies; - if (!rt_caching(dev_net(rt->u.dst.dev))) { + if (!rt_caching(dev_net(rt->dst.dev))) { /* * If we're not caching, just tell the caller we * were successful and don't touch the route. The @@ -1108,7 +1108,7 @@ restart: */ if (rt->rt_type == RTN_UNICAST || rt->fl.iif == 0) { - int err = arp_bind_neighbour(&rt->u.dst); + int err = arp_bind_neighbour(&rt->dst); if (err) { if (net_ratelimit()) printk(KERN_WARNING @@ -1127,19 +1127,19 @@ restart: spin_lock_bh(rt_hash_lock_addr(hash)); while ((rth = *rthp) != NULL) { if (rt_is_expired(rth)) { - *rthp = rth->u.dst.rt_next; + *rthp = rth->dst.rt_next; rt_free(rth); continue; } if (compare_keys(&rth->fl, &rt->fl) && compare_netns(rth, rt)) { /* Put it first */ - *rthp = rth->u.dst.rt_next; + *rthp = rth->dst.rt_next; /* * Since lookup is lockfree, the deletion * must be visible to another weakly ordered CPU before * the insertion at the start of the hash chain. */ - rcu_assign_pointer(rth->u.dst.rt_next, + rcu_assign_pointer(rth->dst.rt_next, rt_hash_table[hash].chain); /* * Since lookup is lockfree, the update writes @@ -1147,18 +1147,18 @@ restart: */ rcu_assign_pointer(rt_hash_table[hash].chain, rth); - dst_use(&rth->u.dst, now); + dst_use(&rth->dst, now); spin_unlock_bh(rt_hash_lock_addr(hash)); rt_drop(rt); if (rp) *rp = rth; else - skb_dst_set(skb, &rth->u.dst); + skb_dst_set(skb, &rth->dst); return 0; } - if (!atomic_read(&rth->u.dst.__refcnt)) { + if (!atomic_read(&rth->dst.__refcnt)) { u32 score = rt_score(rth); if (score <= min_score) { @@ -1170,7 +1170,7 @@ restart: chain_length++; - rthp = &rth->u.dst.rt_next; + rthp = &rth->dst.rt_next; } if (cand) { @@ -1181,17 +1181,17 @@ restart: * only 2 entries per bucket. We will see. */ if (chain_length > ip_rt_gc_elasticity) { - *candp = cand->u.dst.rt_next; + *candp = cand->dst.rt_next; rt_free(cand); } } else { if (chain_length > rt_chain_length_max && slow_chain_length(rt_hash_table[hash].chain) > rt_chain_length_max) { - struct net *net = dev_net(rt->u.dst.dev); + struct net *net = dev_net(rt->dst.dev); int num = ++net->ipv4.current_rt_cache_rebuild_count; if (!rt_caching(net)) { printk(KERN_WARNING "%s: %d rebuilds is over limit, route caching disabled\n", - rt->u.dst.dev->name, num); + rt->dst.dev->name, num); } rt_emergency_hash_rebuild(net); spin_unlock_bh(rt_hash_lock_addr(hash)); @@ -1206,7 +1206,7 @@ restart: route or unicast forwarding path. */ if (rt->rt_type == RTN_UNICAST || rt->fl.iif == 0) { - int err = arp_bind_neighbour(&rt->u.dst); + int err = arp_bind_neighbour(&rt->dst); if (err) { spin_unlock_bh(rt_hash_lock_addr(hash)); @@ -1237,14 +1237,14 @@ restart: } } - rt->u.dst.rt_next = rt_hash_table[hash].chain; + rt->dst.rt_next = rt_hash_table[hash].chain; #if RT_CACHE_DEBUG >= 2 - if (rt->u.dst.rt_next) { + if (rt->dst.rt_next) { struct rtable *trt; printk(KERN_DEBUG "rt_cache @%02x: %pI4", hash, &rt->rt_dst); - for (trt = rt->u.dst.rt_next; trt; trt = trt->u.dst.rt_next) + for (trt = rt->dst.rt_next; trt; trt = trt->dst.rt_next) printk(" . %pI4", &trt->rt_dst); printk("\n"); } @@ -1262,7 +1262,7 @@ skip_hashing: if (rp) *rp = rt; else - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); return 0; } @@ -1334,11 +1334,11 @@ static void rt_del(unsigned hash, struct rtable *rt) ip_rt_put(rt); while ((aux = *rthp) != NULL) { if (aux == rt || rt_is_expired(aux)) { - *rthp = aux->u.dst.rt_next; + *rthp = aux->dst.rt_next; rt_free(aux); continue; } - rthp = &aux->u.dst.rt_next; + rthp = &aux->dst.rt_next; } spin_unlock_bh(rt_hash_lock_addr(hash)); } @@ -1392,19 +1392,19 @@ void ip_rt_redirect(__be32 old_gw, __be32 daddr, __be32 new_gw, rth->fl.oif != ikeys[k] || rth->fl.iif != 0 || rt_is_expired(rth) || - !net_eq(dev_net(rth->u.dst.dev), net)) { - rthp = &rth->u.dst.rt_next; + !net_eq(dev_net(rth->dst.dev), net)) { + rthp = &rth->dst.rt_next; continue; } if (rth->rt_dst != daddr || rth->rt_src != saddr || - rth->u.dst.error || + rth->dst.error || rth->rt_gateway != old_gw || - rth->u.dst.dev != dev) + rth->dst.dev != dev) break; - dst_hold(&rth->u.dst); + dst_hold(&rth->dst); rt = dst_alloc(&ipv4_dst_ops); if (rt == NULL) { @@ -1414,20 +1414,20 @@ void ip_rt_redirect(__be32 old_gw, __be32 daddr, __be32 new_gw, /* Copy all the information. */ *rt = *rth; - rt->u.dst.__use = 1; - atomic_set(&rt->u.dst.__refcnt, 1); - rt->u.dst.child = NULL; - if (rt->u.dst.dev) - dev_hold(rt->u.dst.dev); + rt->dst.__use = 1; + atomic_set(&rt->dst.__refcnt, 1); + rt->dst.child = NULL; + if (rt->dst.dev) + dev_hold(rt->dst.dev); if (rt->idev) in_dev_hold(rt->idev); - rt->u.dst.obsolete = -1; - rt->u.dst.lastuse = jiffies; - rt->u.dst.path = &rt->u.dst; - rt->u.dst.neighbour = NULL; - rt->u.dst.hh = NULL; + rt->dst.obsolete = -1; + rt->dst.lastuse = jiffies; + rt->dst.path = &rt->dst; + rt->dst.neighbour = NULL; + rt->dst.hh = NULL; #ifdef CONFIG_XFRM - rt->u.dst.xfrm = NULL; + rt->dst.xfrm = NULL; #endif rt->rt_genid = rt_genid(net); rt->rt_flags |= RTCF_REDIRECTED; @@ -1436,23 +1436,23 @@ void ip_rt_redirect(__be32 old_gw, __be32 daddr, __be32 new_gw, rt->rt_gateway = new_gw; /* Redirect received -> path was valid */ - dst_confirm(&rth->u.dst); + dst_confirm(&rth->dst); if (rt->peer) atomic_inc(&rt->peer->refcnt); - if (arp_bind_neighbour(&rt->u.dst) || - !(rt->u.dst.neighbour->nud_state & + if (arp_bind_neighbour(&rt->dst) || + !(rt->dst.neighbour->nud_state & NUD_VALID)) { - if (rt->u.dst.neighbour) - neigh_event_send(rt->u.dst.neighbour, NULL); + if (rt->dst.neighbour) + neigh_event_send(rt->dst.neighbour, NULL); ip_rt_put(rth); rt_drop(rt); goto do_next; } - netevent.old = &rth->u.dst; - netevent.new = &rt->u.dst; + netevent.old = &rth->dst; + netevent.new = &rt->dst; call_netevent_notifiers(NETEVENT_REDIRECT, &netevent); @@ -1488,8 +1488,8 @@ static struct dst_entry *ipv4_negative_advice(struct dst_entry *dst) ip_rt_put(rt); ret = NULL; } else if ((rt->rt_flags & RTCF_REDIRECTED) || - (rt->u.dst.expires && - time_after_eq(jiffies, rt->u.dst.expires))) { + (rt->dst.expires && + time_after_eq(jiffies, rt->dst.expires))) { unsigned hash = rt_hash(rt->fl.fl4_dst, rt->fl.fl4_src, rt->fl.oif, rt_genid(dev_net(dst->dev))); @@ -1527,7 +1527,7 @@ void ip_rt_send_redirect(struct sk_buff *skb) int log_martians; rcu_read_lock(); - in_dev = __in_dev_get_rcu(rt->u.dst.dev); + in_dev = __in_dev_get_rcu(rt->dst.dev); if (!in_dev || !IN_DEV_TX_REDIRECTS(in_dev)) { rcu_read_unlock(); return; @@ -1538,30 +1538,30 @@ void ip_rt_send_redirect(struct sk_buff *skb) /* No redirected packets during ip_rt_redirect_silence; * reset the algorithm. */ - if (time_after(jiffies, rt->u.dst.rate_last + ip_rt_redirect_silence)) - rt->u.dst.rate_tokens = 0; + if (time_after(jiffies, rt->dst.rate_last + ip_rt_redirect_silence)) + rt->dst.rate_tokens = 0; /* Too many ignored redirects; do not send anything - * set u.dst.rate_last to the last seen redirected packet. + * set dst.rate_last to the last seen redirected packet. */ - if (rt->u.dst.rate_tokens >= ip_rt_redirect_number) { - rt->u.dst.rate_last = jiffies; + if (rt->dst.rate_tokens >= ip_rt_redirect_number) { + rt->dst.rate_last = jiffies; return; } /* Check for load limit; set rate_last to the latest sent * redirect. */ - if (rt->u.dst.rate_tokens == 0 || + if (rt->dst.rate_tokens == 0 || time_after(jiffies, - (rt->u.dst.rate_last + - (ip_rt_redirect_load << rt->u.dst.rate_tokens)))) { + (rt->dst.rate_last + + (ip_rt_redirect_load << rt->dst.rate_tokens)))) { icmp_send(skb, ICMP_REDIRECT, ICMP_REDIR_HOST, rt->rt_gateway); - rt->u.dst.rate_last = jiffies; - ++rt->u.dst.rate_tokens; + rt->dst.rate_last = jiffies; + ++rt->dst.rate_tokens; #ifdef CONFIG_IP_ROUTE_VERBOSE if (log_martians && - rt->u.dst.rate_tokens == ip_rt_redirect_number && + rt->dst.rate_tokens == ip_rt_redirect_number && net_ratelimit()) printk(KERN_WARNING "host %pI4/if%d ignores redirects for %pI4 to %pI4.\n", &rt->rt_src, rt->rt_iif, @@ -1576,7 +1576,7 @@ static int ip_error(struct sk_buff *skb) unsigned long now; int code; - switch (rt->u.dst.error) { + switch (rt->dst.error) { case EINVAL: default: goto out; @@ -1585,7 +1585,7 @@ static int ip_error(struct sk_buff *skb) break; case ENETUNREACH: code = ICMP_NET_UNREACH; - IP_INC_STATS_BH(dev_net(rt->u.dst.dev), + IP_INC_STATS_BH(dev_net(rt->dst.dev), IPSTATS_MIB_INNOROUTES); break; case EACCES: @@ -1594,12 +1594,12 @@ static int ip_error(struct sk_buff *skb) } now = jiffies; - rt->u.dst.rate_tokens += now - rt->u.dst.rate_last; - if (rt->u.dst.rate_tokens > ip_rt_error_burst) - rt->u.dst.rate_tokens = ip_rt_error_burst; - rt->u.dst.rate_last = now; - if (rt->u.dst.rate_tokens >= ip_rt_error_cost) { - rt->u.dst.rate_tokens -= ip_rt_error_cost; + rt->dst.rate_tokens += now - rt->dst.rate_last; + if (rt->dst.rate_tokens > ip_rt_error_burst) + rt->dst.rate_tokens = ip_rt_error_burst; + rt->dst.rate_last = now; + if (rt->dst.rate_tokens >= ip_rt_error_cost) { + rt->dst.rate_tokens -= ip_rt_error_cost; icmp_send(skb, ICMP_DEST_UNREACH, code, 0); } @@ -1644,7 +1644,7 @@ unsigned short ip_rt_frag_needed(struct net *net, struct iphdr *iph, rcu_read_lock(); for (rth = rcu_dereference(rt_hash_table[hash].chain); rth; - rth = rcu_dereference(rth->u.dst.rt_next)) { + rth = rcu_dereference(rth->dst.rt_next)) { unsigned short mtu = new_mtu; if (rth->fl.fl4_dst != daddr || @@ -1653,8 +1653,8 @@ unsigned short ip_rt_frag_needed(struct net *net, struct iphdr *iph, rth->rt_src != iph->saddr || rth->fl.oif != ikeys[k] || rth->fl.iif != 0 || - dst_metric_locked(&rth->u.dst, RTAX_MTU) || - !net_eq(dev_net(rth->u.dst.dev), net) || + dst_metric_locked(&rth->dst, RTAX_MTU) || + !net_eq(dev_net(rth->dst.dev), net) || rt_is_expired(rth)) continue; @@ -1662,22 +1662,22 @@ unsigned short ip_rt_frag_needed(struct net *net, struct iphdr *iph, /* BSD 4.2 compatibility hack :-( */ if (mtu == 0 && - old_mtu >= dst_mtu(&rth->u.dst) && + old_mtu >= dst_mtu(&rth->dst) && old_mtu >= 68 + (iph->ihl << 2)) old_mtu -= iph->ihl << 2; mtu = guess_mtu(old_mtu); } - if (mtu <= dst_mtu(&rth->u.dst)) { - if (mtu < dst_mtu(&rth->u.dst)) { - dst_confirm(&rth->u.dst); + if (mtu <= dst_mtu(&rth->dst)) { + if (mtu < dst_mtu(&rth->dst)) { + dst_confirm(&rth->dst); if (mtu < ip_rt_min_pmtu) { mtu = ip_rt_min_pmtu; - rth->u.dst.metrics[RTAX_LOCK-1] |= + rth->dst.metrics[RTAX_LOCK-1] |= (1 << RTAX_MTU); } - rth->u.dst.metrics[RTAX_MTU-1] = mtu; - dst_set_expires(&rth->u.dst, + rth->dst.metrics[RTAX_MTU-1] = mtu; + dst_set_expires(&rth->dst, ip_rt_mtu_expires); } est_mtu = mtu; @@ -1750,7 +1750,7 @@ static void ipv4_link_failure(struct sk_buff *skb) rt = skb_rtable(skb); if (rt) - dst_set_expires(&rt->u.dst, 0); + dst_set_expires(&rt->dst, 0); } static int ip_rt_bug(struct sk_buff *skb) @@ -1778,11 +1778,11 @@ void ip_rt_get_source(u8 *addr, struct rtable *rt) if (rt->fl.iif == 0) src = rt->rt_src; - else if (fib_lookup(dev_net(rt->u.dst.dev), &rt->fl, &res) == 0) { + else if (fib_lookup(dev_net(rt->dst.dev), &rt->fl, &res) == 0) { src = FIB_RES_PREFSRC(res); fib_res_put(&res); } else - src = inet_select_addr(rt->u.dst.dev, rt->rt_gateway, + src = inet_select_addr(rt->dst.dev, rt->rt_gateway, RT_SCOPE_UNIVERSE); memcpy(addr, &src, 4); } @@ -1790,10 +1790,10 @@ void ip_rt_get_source(u8 *addr, struct rtable *rt) #ifdef CONFIG_NET_CLS_ROUTE static void set_class_tag(struct rtable *rt, u32 tag) { - if (!(rt->u.dst.tclassid & 0xFFFF)) - rt->u.dst.tclassid |= tag & 0xFFFF; - if (!(rt->u.dst.tclassid & 0xFFFF0000)) - rt->u.dst.tclassid |= tag & 0xFFFF0000; + if (!(rt->dst.tclassid & 0xFFFF)) + rt->dst.tclassid |= tag & 0xFFFF; + if (!(rt->dst.tclassid & 0xFFFF0000)) + rt->dst.tclassid |= tag & 0xFFFF0000; } #endif @@ -1805,30 +1805,30 @@ static void rt_set_nexthop(struct rtable *rt, struct fib_result *res, u32 itag) if (FIB_RES_GW(*res) && FIB_RES_NH(*res).nh_scope == RT_SCOPE_LINK) rt->rt_gateway = FIB_RES_GW(*res); - memcpy(rt->u.dst.metrics, fi->fib_metrics, - sizeof(rt->u.dst.metrics)); + memcpy(rt->dst.metrics, fi->fib_metrics, + sizeof(rt->dst.metrics)); if (fi->fib_mtu == 0) { - rt->u.dst.metrics[RTAX_MTU-1] = rt->u.dst.dev->mtu; - if (dst_metric_locked(&rt->u.dst, RTAX_MTU) && + rt->dst.metrics[RTAX_MTU-1] = rt->dst.dev->mtu; + if (dst_metric_locked(&rt->dst, RTAX_MTU) && rt->rt_gateway != rt->rt_dst && - rt->u.dst.dev->mtu > 576) - rt->u.dst.metrics[RTAX_MTU-1] = 576; + rt->dst.dev->mtu > 576) + rt->dst.metrics[RTAX_MTU-1] = 576; } #ifdef CONFIG_NET_CLS_ROUTE - rt->u.dst.tclassid = FIB_RES_NH(*res).nh_tclassid; + rt->dst.tclassid = FIB_RES_NH(*res).nh_tclassid; #endif } else - rt->u.dst.metrics[RTAX_MTU-1]= rt->u.dst.dev->mtu; - - if (dst_metric(&rt->u.dst, RTAX_HOPLIMIT) == 0) - rt->u.dst.metrics[RTAX_HOPLIMIT-1] = sysctl_ip_default_ttl; - if (dst_mtu(&rt->u.dst) > IP_MAX_MTU) - rt->u.dst.metrics[RTAX_MTU-1] = IP_MAX_MTU; - if (dst_metric(&rt->u.dst, RTAX_ADVMSS) == 0) - rt->u.dst.metrics[RTAX_ADVMSS-1] = max_t(unsigned int, rt->u.dst.dev->mtu - 40, + rt->dst.metrics[RTAX_MTU-1]= rt->dst.dev->mtu; + + if (dst_metric(&rt->dst, RTAX_HOPLIMIT) == 0) + rt->dst.metrics[RTAX_HOPLIMIT-1] = sysctl_ip_default_ttl; + if (dst_mtu(&rt->dst) > IP_MAX_MTU) + rt->dst.metrics[RTAX_MTU-1] = IP_MAX_MTU; + if (dst_metric(&rt->dst, RTAX_ADVMSS) == 0) + rt->dst.metrics[RTAX_ADVMSS-1] = max_t(unsigned int, rt->dst.dev->mtu - 40, ip_rt_min_advmss); - if (dst_metric(&rt->u.dst, RTAX_ADVMSS) > 65535 - 40) - rt->u.dst.metrics[RTAX_ADVMSS-1] = 65535 - 40; + if (dst_metric(&rt->dst, RTAX_ADVMSS) > 65535 - 40) + rt->dst.metrics[RTAX_ADVMSS-1] = 65535 - 40; #ifdef CONFIG_NET_CLS_ROUTE #ifdef CONFIG_IP_MULTIPLE_TABLES @@ -1873,13 +1873,13 @@ static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr, if (!rth) goto e_nobufs; - rth->u.dst.output = ip_rt_bug; - rth->u.dst.obsolete = -1; + rth->dst.output = ip_rt_bug; + rth->dst.obsolete = -1; - atomic_set(&rth->u.dst.__refcnt, 1); - rth->u.dst.flags= DST_HOST; + atomic_set(&rth->dst.__refcnt, 1); + rth->dst.flags= DST_HOST; if (IN_DEV_CONF_GET(in_dev, NOPOLICY)) - rth->u.dst.flags |= DST_NOPOLICY; + rth->dst.flags |= DST_NOPOLICY; rth->fl.fl4_dst = daddr; rth->rt_dst = daddr; rth->fl.fl4_tos = tos; @@ -1887,13 +1887,13 @@ static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr, rth->fl.fl4_src = saddr; rth->rt_src = saddr; #ifdef CONFIG_NET_CLS_ROUTE - rth->u.dst.tclassid = itag; + rth->dst.tclassid = itag; #endif rth->rt_iif = rth->fl.iif = dev->ifindex; - rth->u.dst.dev = init_net.loopback_dev; - dev_hold(rth->u.dst.dev); - rth->idev = in_dev_get(rth->u.dst.dev); + rth->dst.dev = init_net.loopback_dev; + dev_hold(rth->dst.dev); + rth->idev = in_dev_get(rth->dst.dev); rth->fl.oif = 0; rth->rt_gateway = daddr; rth->rt_spec_dst= spec_dst; @@ -1901,13 +1901,13 @@ static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr, rth->rt_flags = RTCF_MULTICAST; rth->rt_type = RTN_MULTICAST; if (our) { - rth->u.dst.input= ip_local_deliver; + rth->dst.input= ip_local_deliver; rth->rt_flags |= RTCF_LOCAL; } #ifdef CONFIG_IP_MROUTE if (!ipv4_is_local_multicast(daddr) && IN_DEV_MFORWARD(in_dev)) - rth->u.dst.input = ip_mr_input; + rth->dst.input = ip_mr_input; #endif RT_CACHE_STAT_INC(in_slow_mc); @@ -2016,12 +2016,12 @@ static int __mkroute_input(struct sk_buff *skb, goto cleanup; } - atomic_set(&rth->u.dst.__refcnt, 1); - rth->u.dst.flags= DST_HOST; + atomic_set(&rth->dst.__refcnt, 1); + rth->dst.flags= DST_HOST; if (IN_DEV_CONF_GET(in_dev, NOPOLICY)) - rth->u.dst.flags |= DST_NOPOLICY; + rth->dst.flags |= DST_NOPOLICY; if (IN_DEV_CONF_GET(out_dev, NOXFRM)) - rth->u.dst.flags |= DST_NOXFRM; + rth->dst.flags |= DST_NOXFRM; rth->fl.fl4_dst = daddr; rth->rt_dst = daddr; rth->fl.fl4_tos = tos; @@ -2031,16 +2031,16 @@ static int __mkroute_input(struct sk_buff *skb, rth->rt_gateway = daddr; rth->rt_iif = rth->fl.iif = in_dev->dev->ifindex; - rth->u.dst.dev = (out_dev)->dev; - dev_hold(rth->u.dst.dev); - rth->idev = in_dev_get(rth->u.dst.dev); + rth->dst.dev = (out_dev)->dev; + dev_hold(rth->dst.dev); + rth->idev = in_dev_get(rth->dst.dev); rth->fl.oif = 0; rth->rt_spec_dst= spec_dst; - rth->u.dst.obsolete = -1; - rth->u.dst.input = ip_forward; - rth->u.dst.output = ip_output; - rth->rt_genid = rt_genid(dev_net(rth->u.dst.dev)); + rth->dst.obsolete = -1; + rth->dst.input = ip_forward; + rth->dst.output = ip_output; + rth->rt_genid = rt_genid(dev_net(rth->dst.dev)); rt_set_nexthop(rth, res, itag); @@ -2074,7 +2074,7 @@ static int ip_mkroute_input(struct sk_buff *skb, /* put it into the cache */ hash = rt_hash(daddr, saddr, fl->iif, - rt_genid(dev_net(rth->u.dst.dev))); + rt_genid(dev_net(rth->dst.dev))); return rt_intern_hash(hash, rth, NULL, skb, fl->iif); } @@ -2197,14 +2197,14 @@ local_input: if (!rth) goto e_nobufs; - rth->u.dst.output= ip_rt_bug; - rth->u.dst.obsolete = -1; + rth->dst.output= ip_rt_bug; + rth->dst.obsolete = -1; rth->rt_genid = rt_genid(net); - atomic_set(&rth->u.dst.__refcnt, 1); - rth->u.dst.flags= DST_HOST; + atomic_set(&rth->dst.__refcnt, 1); + rth->dst.flags= DST_HOST; if (IN_DEV_CONF_GET(in_dev, NOPOLICY)) - rth->u.dst.flags |= DST_NOPOLICY; + rth->dst.flags |= DST_NOPOLICY; rth->fl.fl4_dst = daddr; rth->rt_dst = daddr; rth->fl.fl4_tos = tos; @@ -2212,20 +2212,20 @@ local_input: rth->fl.fl4_src = saddr; rth->rt_src = saddr; #ifdef CONFIG_NET_CLS_ROUTE - rth->u.dst.tclassid = itag; + rth->dst.tclassid = itag; #endif rth->rt_iif = rth->fl.iif = dev->ifindex; - rth->u.dst.dev = net->loopback_dev; - dev_hold(rth->u.dst.dev); - rth->idev = in_dev_get(rth->u.dst.dev); + rth->dst.dev = net->loopback_dev; + dev_hold(rth->dst.dev); + rth->idev = in_dev_get(rth->dst.dev); rth->rt_gateway = daddr; rth->rt_spec_dst= spec_dst; - rth->u.dst.input= ip_local_deliver; + rth->dst.input= ip_local_deliver; rth->rt_flags = flags|RTCF_LOCAL; if (res.type == RTN_UNREACHABLE) { - rth->u.dst.input= ip_error; - rth->u.dst.error= -err; + rth->dst.input= ip_error; + rth->dst.error= -err; rth->rt_flags &= ~RTCF_LOCAL; } rth->rt_type = res.type; @@ -2291,21 +2291,21 @@ int ip_route_input_common(struct sk_buff *skb, __be32 daddr, __be32 saddr, hash = rt_hash(daddr, saddr, iif, rt_genid(net)); for (rth = rcu_dereference(rt_hash_table[hash].chain); rth; - rth = rcu_dereference(rth->u.dst.rt_next)) { + rth = rcu_dereference(rth->dst.rt_next)) { if ((((__force u32)rth->fl.fl4_dst ^ (__force u32)daddr) | ((__force u32)rth->fl.fl4_src ^ (__force u32)saddr) | (rth->fl.iif ^ iif) | rth->fl.oif | (rth->fl.fl4_tos ^ tos)) == 0 && rth->fl.mark == skb->mark && - net_eq(dev_net(rth->u.dst.dev), net) && + net_eq(dev_net(rth->dst.dev), net) && !rt_is_expired(rth)) { if (noref) { - dst_use_noref(&rth->u.dst, jiffies); - skb_dst_set_noref(skb, &rth->u.dst); + dst_use_noref(&rth->dst, jiffies); + skb_dst_set_noref(skb, &rth->dst); } else { - dst_use(&rth->u.dst, jiffies); - skb_dst_set(skb, &rth->u.dst); + dst_use(&rth->dst, jiffies); + skb_dst_set(skb, &rth->dst); } RT_CACHE_STAT_INC(in_hit); rcu_read_unlock(); @@ -2412,12 +2412,12 @@ static int __mkroute_output(struct rtable **result, goto cleanup; } - atomic_set(&rth->u.dst.__refcnt, 1); - rth->u.dst.flags= DST_HOST; + atomic_set(&rth->dst.__refcnt, 1); + rth->dst.flags= DST_HOST; if (IN_DEV_CONF_GET(in_dev, NOXFRM)) - rth->u.dst.flags |= DST_NOXFRM; + rth->dst.flags |= DST_NOXFRM; if (IN_DEV_CONF_GET(in_dev, NOPOLICY)) - rth->u.dst.flags |= DST_NOPOLICY; + rth->dst.flags |= DST_NOPOLICY; rth->fl.fl4_dst = oldflp->fl4_dst; rth->fl.fl4_tos = tos; @@ -2429,35 +2429,35 @@ static int __mkroute_output(struct rtable **result, rth->rt_iif = oldflp->oif ? : dev_out->ifindex; /* get references to the devices that are to be hold by the routing cache entry */ - rth->u.dst.dev = dev_out; + rth->dst.dev = dev_out; dev_hold(dev_out); rth->idev = in_dev_get(dev_out); rth->rt_gateway = fl->fl4_dst; rth->rt_spec_dst= fl->fl4_src; - rth->u.dst.output=ip_output; - rth->u.dst.obsolete = -1; + rth->dst.output=ip_output; + rth->dst.obsolete = -1; rth->rt_genid = rt_genid(dev_net(dev_out)); RT_CACHE_STAT_INC(out_slow_tot); if (flags & RTCF_LOCAL) { - rth->u.dst.input = ip_local_deliver; + rth->dst.input = ip_local_deliver; rth->rt_spec_dst = fl->fl4_dst; } if (flags & (RTCF_BROADCAST | RTCF_MULTICAST)) { rth->rt_spec_dst = fl->fl4_src; if (flags & RTCF_LOCAL && !(dev_out->flags & IFF_LOOPBACK)) { - rth->u.dst.output = ip_mc_output; + rth->dst.output = ip_mc_output; RT_CACHE_STAT_INC(out_slow_mc); } #ifdef CONFIG_IP_MROUTE if (res->type == RTN_MULTICAST) { if (IN_DEV_MFORWARD(in_dev) && !ipv4_is_local_multicast(oldflp->fl4_dst)) { - rth->u.dst.input = ip_mr_input; - rth->u.dst.output = ip_mc_output; + rth->dst.input = ip_mr_input; + rth->dst.output = ip_mc_output; } } #endif @@ -2712,7 +2712,7 @@ int __ip_route_output_key(struct net *net, struct rtable **rp, rcu_read_lock_bh(); for (rth = rcu_dereference_bh(rt_hash_table[hash].chain); rth; - rth = rcu_dereference_bh(rth->u.dst.rt_next)) { + rth = rcu_dereference_bh(rth->dst.rt_next)) { if (rth->fl.fl4_dst == flp->fl4_dst && rth->fl.fl4_src == flp->fl4_src && rth->fl.iif == 0 && @@ -2720,9 +2720,9 @@ int __ip_route_output_key(struct net *net, struct rtable **rp, rth->fl.mark == flp->mark && !((rth->fl.fl4_tos ^ flp->fl4_tos) & (IPTOS_RT_MASK | RTO_ONLINK)) && - net_eq(dev_net(rth->u.dst.dev), net) && + net_eq(dev_net(rth->dst.dev), net) && !rt_is_expired(rth)) { - dst_use(&rth->u.dst, jiffies); + dst_use(&rth->dst, jiffies); RT_CACHE_STAT_INC(out_hit); rcu_read_unlock_bh(); *rp = rth; @@ -2759,15 +2759,15 @@ static int ipv4_dst_blackhole(struct net *net, struct rtable **rp, struct flowi dst_alloc(&ipv4_dst_blackhole_ops); if (rt) { - struct dst_entry *new = &rt->u.dst; + struct dst_entry *new = &rt->dst; atomic_set(&new->__refcnt, 1); new->__use = 1; new->input = dst_discard; new->output = dst_discard; - memcpy(new->metrics, ort->u.dst.metrics, RTAX_MAX*sizeof(u32)); + memcpy(new->metrics, ort->dst.metrics, RTAX_MAX*sizeof(u32)); - new->dev = ort->u.dst.dev; + new->dev = ort->dst.dev; if (new->dev) dev_hold(new->dev); @@ -2791,7 +2791,7 @@ static int ipv4_dst_blackhole(struct net *net, struct rtable **rp, struct flowi dst_free(new); } - dst_release(&(*rp)->u.dst); + dst_release(&(*rp)->dst); *rp = rt; return (rt ? 0 : -ENOMEM); } @@ -2861,11 +2861,11 @@ static int rt_fill_info(struct net *net, r->rtm_src_len = 32; NLA_PUT_BE32(skb, RTA_SRC, rt->fl.fl4_src); } - if (rt->u.dst.dev) - NLA_PUT_U32(skb, RTA_OIF, rt->u.dst.dev->ifindex); + if (rt->dst.dev) + NLA_PUT_U32(skb, RTA_OIF, rt->dst.dev->ifindex); #ifdef CONFIG_NET_CLS_ROUTE - if (rt->u.dst.tclassid) - NLA_PUT_U32(skb, RTA_FLOW, rt->u.dst.tclassid); + if (rt->dst.tclassid) + NLA_PUT_U32(skb, RTA_FLOW, rt->dst.tclassid); #endif if (rt->fl.iif) NLA_PUT_BE32(skb, RTA_PREFSRC, rt->rt_spec_dst); @@ -2875,11 +2875,11 @@ static int rt_fill_info(struct net *net, if (rt->rt_dst != rt->rt_gateway) NLA_PUT_BE32(skb, RTA_GATEWAY, rt->rt_gateway); - if (rtnetlink_put_metrics(skb, rt->u.dst.metrics) < 0) + if (rtnetlink_put_metrics(skb, rt->dst.metrics) < 0) goto nla_put_failure; - error = rt->u.dst.error; - expires = rt->u.dst.expires ? rt->u.dst.expires - jiffies : 0; + error = rt->dst.error; + expires = rt->dst.expires ? rt->dst.expires - jiffies : 0; if (rt->peer) { id = atomic_read(&rt->peer->ip_id_count) & 0xffff; if (rt->peer->tcp_ts_stamp) { @@ -2911,7 +2911,7 @@ static int rt_fill_info(struct net *net, NLA_PUT_U32(skb, RTA_IIF, rt->fl.iif); } - if (rtnl_put_cacheinfo(skb, &rt->u.dst, id, ts, tsage, + if (rtnl_put_cacheinfo(skb, &rt->dst, id, ts, tsage, expires, error) < 0) goto nla_put_failure; @@ -2976,8 +2976,8 @@ static int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void local_bh_enable(); rt = skb_rtable(skb); - if (err == 0 && rt->u.dst.error) - err = -rt->u.dst.error; + if (err == 0 && rt->dst.error) + err = -rt->dst.error; } else { struct flowi fl = { .nl_u = { @@ -2995,7 +2995,7 @@ static int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void if (err) goto errout_free; - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); if (rtm->rtm_flags & RTM_F_NOTIFY) rt->rt_flags |= RTCF_NOTIFY; @@ -3031,12 +3031,12 @@ int ip_rt_dump(struct sk_buff *skb, struct netlink_callback *cb) continue; rcu_read_lock_bh(); for (rt = rcu_dereference_bh(rt_hash_table[h].chain), idx = 0; rt; - rt = rcu_dereference_bh(rt->u.dst.rt_next), idx++) { - if (!net_eq(dev_net(rt->u.dst.dev), net) || idx < s_idx) + rt = rcu_dereference_bh(rt->dst.rt_next), idx++) { + if (!net_eq(dev_net(rt->dst.dev), net) || idx < s_idx) continue; if (rt_is_expired(rt)) continue; - skb_dst_set_noref(skb, &rt->u.dst); + skb_dst_set_noref(skb, &rt->dst); if (rt_fill_info(net, skb, NETLINK_CB(cb->skb).pid, cb->nlh->nlmsg_seq, RTM_NEWROUTE, 1, NLM_F_MULTI) <= 0) { diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c index 5c48124..02bef6a 100644 --- a/net/ipv4/syncookies.c +++ b/net/ipv4/syncookies.c @@ -354,15 +354,15 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb, } /* Try to redo what tcp_v4_send_synack did. */ - req->window_clamp = tp->window_clamp ? :dst_metric(&rt->u.dst, RTAX_WINDOW); + req->window_clamp = tp->window_clamp ? :dst_metric(&rt->dst, RTAX_WINDOW); tcp_select_initial_window(tcp_full_space(sk), req->mss, &req->rcv_wnd, &req->window_clamp, ireq->wscale_ok, &rcv_wscale, - dst_metric(&rt->u.dst, RTAX_INITRWND)); + dst_metric(&rt->dst, RTAX_INITRWND)); ireq->rcv_wscale = rcv_wscale; - ret = get_cookie_sock(sk, skb, req, &rt->u.dst); + ret = get_cookie_sock(sk, skb, req, &rt->dst); out: return ret; } diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 7f976af..7f9515c 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -237,7 +237,7 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) /* OK, now commit destination to socket. */ sk->sk_gso_type = SKB_GSO_TCPV4; - sk_setup_caps(sk, &rt->u.dst); + sk_setup_caps(sk, &rt->dst); if (!tp->write_seq) tp->write_seq = secure_tcp_sequence_number(inet->inet_saddr, diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index eec4ff4..32e0bef 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -914,7 +914,7 @@ int udp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, !sock_flag(sk, SOCK_BROADCAST)) goto out; if (connected) - sk_dst_set(sk, dst_clone(&rt->u.dst)); + sk_dst_set(sk, dst_clone(&rt->dst)); } if (msg->msg_flags&MSG_CONFIRM) @@ -978,7 +978,7 @@ out: return err; do_confirm: - dst_confirm(&rt->u.dst); + dst_confirm(&rt->dst); if (!(msg->msg_flags&MSG_PROBE) || len) goto back_from_confirm; err = 0; diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c index 1705476..3493270 100644 --- a/net/ipv4/xfrm4_policy.c +++ b/net/ipv4/xfrm4_policy.c @@ -37,7 +37,7 @@ static struct dst_entry *xfrm4_dst_lookup(struct net *net, int tos, fl.fl4_src = saddr->a4; err = __ip_route_output_key(net, &rt, &fl); - dst = &rt->u.dst; + dst = &rt->dst; if (err) dst = ERR_PTR(err); return dst; diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index e1a698d..b97bb1f 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -557,7 +557,7 @@ void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp) pr_warning("Freeing alive inet6 address %p\n", ifp); return; } - dst_release(&ifp->rt->u.dst); + dst_release(&ifp->rt->dst); call_rcu(&ifp->rcu, inet6_ifa_finish_destroy_rcu); } @@ -823,7 +823,7 @@ static void ipv6_del_addr(struct inet6_ifaddr *ifp) rt->rt6i_flags |= RTF_EXPIRES; } } - dst_release(&rt->u.dst); + dst_release(&rt->dst); } out: @@ -1863,7 +1863,7 @@ void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len) dev, expires, flags); } if (rt) - dst_release(&rt->u.dst); + dst_release(&rt->dst); } /* Try to figure out our local address for this prefix */ @@ -4093,11 +4093,11 @@ static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp) if (ifp->idev->cnf.forwarding) addrconf_leave_anycast(ifp); addrconf_leave_solict(ifp->idev, &ifp->addr); - dst_hold(&ifp->rt->u.dst); + dst_hold(&ifp->rt->dst); if (ifp->state == INET6_IFADDR_STATE_DEAD && ip6_del_rt(ifp->rt)) - dst_free(&ifp->rt->u.dst); + dst_free(&ifp->rt->dst); break; } } diff --git a/net/ipv6/anycast.c b/net/ipv6/anycast.c index f058fbd..0e5e943 100644 --- a/net/ipv6/anycast.c +++ b/net/ipv6/anycast.c @@ -84,7 +84,7 @@ int ipv6_sock_ac_join(struct sock *sk, int ifindex, struct in6_addr *addr) rt = rt6_lookup(net, addr, NULL, 0, 0); if (rt) { dev = rt->rt6i_dev; - dst_release(&rt->u.dst); + dst_release(&rt->dst); } else if (ishost) { err = -EADDRNOTAVAIL; goto error; @@ -244,7 +244,7 @@ static void aca_put(struct ifacaddr6 *ac) { if (atomic_dec_and_test(&ac->aca_refcnt)) { in6_dev_put(ac->aca_idev); - dst_release(&ac->aca_rt->u.dst); + dst_release(&ac->aca_rt->dst); kfree(ac); } } @@ -350,7 +350,7 @@ int __ipv6_dev_ac_dec(struct inet6_dev *idev, struct in6_addr *addr) write_unlock_bh(&idev->lock); addrconf_leave_solict(idev, &aca->aca_addr); - dst_hold(&aca->aca_rt->u.dst); + dst_hold(&aca->aca_rt->dst); ip6_del_rt(aca->aca_rt); aca_put(aca); diff --git a/net/ipv6/fib6_rules.c b/net/ipv6/fib6_rules.c index 8e44f8f..b1108ed 100644 --- a/net/ipv6/fib6_rules.c +++ b/net/ipv6/fib6_rules.c @@ -43,8 +43,8 @@ struct dst_entry *fib6_rule_lookup(struct net *net, struct flowi *fl, if (arg.result) return arg.result; - dst_hold(&net->ipv6.ip6_null_entry->u.dst); - return &net->ipv6.ip6_null_entry->u.dst; + dst_hold(&net->ipv6.ip6_null_entry->dst); + return &net->ipv6.ip6_null_entry->dst; } static int fib6_rule_action(struct fib_rule *rule, struct flowi *flp, @@ -86,7 +86,7 @@ static int fib6_rule_action(struct fib_rule *rule, struct flowi *flp, struct in6_addr saddr; if (ipv6_dev_get_saddr(net, - ip6_dst_idev(&rt->u.dst)->dev, + ip6_dst_idev(&rt->dst)->dev, &flp->fl6_dst, rt6_flags2srcprefs(flags), &saddr)) @@ -99,12 +99,12 @@ static int fib6_rule_action(struct fib_rule *rule, struct flowi *flp, goto out; } again: - dst_release(&rt->u.dst); + dst_release(&rt->dst); rt = NULL; goto out; discard_pkt: - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); out: arg->result = rt; return rt == NULL ? -EAGAIN : 0; diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c index 92a122b..b6a5859 100644 --- a/net/ipv6/ip6_fib.c +++ b/net/ipv6/ip6_fib.c @@ -165,7 +165,7 @@ static __inline__ void node_free(struct fib6_node * fn) static __inline__ void rt6_release(struct rt6_info *rt) { if (atomic_dec_and_test(&rt->rt6i_ref)) - dst_free(&rt->u.dst); + dst_free(&rt->dst); } static void fib6_link_table(struct net *net, struct fib6_table *tb) @@ -278,7 +278,7 @@ static int fib6_dump_node(struct fib6_walker_t *w) int res; struct rt6_info *rt; - for (rt = w->leaf; rt; rt = rt->u.dst.rt6_next) { + for (rt = w->leaf; rt; rt = rt->dst.rt6_next) { res = rt6_dump_route(rt, w->args); if (res < 0) { /* Frame is full, suspend walking */ @@ -619,7 +619,7 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt, ins = &fn->leaf; - for (iter = fn->leaf; iter; iter=iter->u.dst.rt6_next) { + for (iter = fn->leaf; iter; iter=iter->dst.rt6_next) { /* * Search for duplicates */ @@ -647,7 +647,7 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt, if (iter->rt6i_metric > rt->rt6i_metric) break; - ins = &iter->u.dst.rt6_next; + ins = &iter->dst.rt6_next; } /* Reset round-robin state, if necessary */ @@ -658,7 +658,7 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt, * insert node */ - rt->u.dst.rt6_next = iter; + rt->dst.rt6_next = iter; *ins = rt; rt->rt6i_node = fn; atomic_inc(&rt->rt6i_ref); @@ -799,7 +799,7 @@ out: atomic_inc(&pn->leaf->rt6i_ref); } #endif - dst_free(&rt->u.dst); + dst_free(&rt->dst); } return err; @@ -810,7 +810,7 @@ out: st_failure: if (fn && !(fn->fn_flags & (RTN_RTINFO|RTN_ROOT))) fib6_repair_tree(info->nl_net, fn); - dst_free(&rt->u.dst); + dst_free(&rt->dst); return err; #endif } @@ -1108,7 +1108,7 @@ static void fib6_del_route(struct fib6_node *fn, struct rt6_info **rtp, RT6_TRACE("fib6_del_route\n"); /* Unlink it */ - *rtp = rt->u.dst.rt6_next; + *rtp = rt->dst.rt6_next; rt->rt6i_node = NULL; net->ipv6.rt6_stats->fib_rt_entries--; net->ipv6.rt6_stats->fib_discarded_routes++; @@ -1122,14 +1122,14 @@ static void fib6_del_route(struct fib6_node *fn, struct rt6_info **rtp, FOR_WALKERS(w) { if (w->state == FWS_C && w->leaf == rt) { RT6_TRACE("walker %p adjusted by delroute\n", w); - w->leaf = rt->u.dst.rt6_next; + w->leaf = rt->dst.rt6_next; if (w->leaf == NULL) w->state = FWS_U; } } read_unlock(&fib6_walker_lock); - rt->u.dst.rt6_next = NULL; + rt->dst.rt6_next = NULL; /* If it was last route, expunge its radix tree node */ if (fn->leaf == NULL) { @@ -1168,7 +1168,7 @@ int fib6_del(struct rt6_info *rt, struct nl_info *info) struct rt6_info **rtp; #if RT6_DEBUG >= 2 - if (rt->u.dst.obsolete>0) { + if (rt->dst.obsolete>0) { WARN_ON(fn != NULL); return -ENOENT; } @@ -1195,7 +1195,7 @@ int fib6_del(struct rt6_info *rt, struct nl_info *info) * Walk the leaf entries looking for ourself */ - for (rtp = &fn->leaf; *rtp; rtp = &(*rtp)->u.dst.rt6_next) { + for (rtp = &fn->leaf; *rtp; rtp = &(*rtp)->dst.rt6_next) { if (*rtp == rt) { fib6_del_route(fn, rtp, info); return 0; @@ -1334,7 +1334,7 @@ static int fib6_clean_node(struct fib6_walker_t *w) .nl_net = c->net, }; - for (rt = w->leaf; rt; rt = rt->u.dst.rt6_next) { + for (rt = w->leaf; rt; rt = rt->dst.rt6_next) { res = c->func(rt, c->arg); if (res < 0) { w->leaf = rt; @@ -1448,8 +1448,8 @@ static int fib6_age(struct rt6_info *rt, void *arg) } gc_args.more++; } else if (rt->rt6i_flags & RTF_CACHE) { - if (atomic_read(&rt->u.dst.__refcnt) == 0 && - time_after_eq(now, rt->u.dst.lastuse + gc_args.timeout)) { + if (atomic_read(&rt->dst.__refcnt) == 0 && + time_after_eq(now, rt->dst.lastuse + gc_args.timeout)) { RT6_TRACE("aging clone %p\n", rt); return -1; } else if ((rt->rt6i_flags & RTF_GATEWAY) && diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 89425af..d40b330 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -698,7 +698,7 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) ipv6_hdr(skb)->payload_len = htons(first_len - sizeof(struct ipv6hdr)); - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); for (;;) { /* Prepare header of the next frame, @@ -726,7 +726,7 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) err = output(skb); if(!err) - IP6_INC_STATS(net, ip6_dst_idev(&rt->u.dst), + IP6_INC_STATS(net, ip6_dst_idev(&rt->dst), IPSTATS_MIB_FRAGCREATES); if (err || !frag) @@ -740,9 +740,9 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) kfree(tmp_hdr); if (err == 0) { - IP6_INC_STATS(net, ip6_dst_idev(&rt->u.dst), + IP6_INC_STATS(net, ip6_dst_idev(&rt->dst), IPSTATS_MIB_FRAGOKS); - dst_release(&rt->u.dst); + dst_release(&rt->dst); return 0; } @@ -752,9 +752,9 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) frag = skb; } - IP6_INC_STATS(net, ip6_dst_idev(&rt->u.dst), + IP6_INC_STATS(net, ip6_dst_idev(&rt->dst), IPSTATS_MIB_FRAGFAILS); - dst_release(&rt->u.dst); + dst_release(&rt->dst); return err; } @@ -785,7 +785,7 @@ slow_path: * Allocate buffer. */ - if ((frag = alloc_skb(len+hlen+sizeof(struct frag_hdr)+LL_ALLOCATED_SPACE(rt->u.dst.dev), GFP_ATOMIC)) == NULL) { + if ((frag = alloc_skb(len+hlen+sizeof(struct frag_hdr)+LL_ALLOCATED_SPACE(rt->dst.dev), GFP_ATOMIC)) == NULL) { NETDEBUG(KERN_INFO "IPv6: frag: no memory for new fragment!\n"); IP6_INC_STATS(net, ip6_dst_idev(skb_dst(skb)), IPSTATS_MIB_FRAGFAILS); @@ -798,7 +798,7 @@ slow_path: */ ip6_copy_metadata(frag, skb); - skb_reserve(frag, LL_RESERVED_SPACE(rt->u.dst.dev)); + skb_reserve(frag, LL_RESERVED_SPACE(rt->dst.dev)); skb_put(frag, len + hlen + sizeof(struct frag_hdr)); skb_reset_network_header(frag); fh = (struct frag_hdr *)(skb_network_header(frag) + hlen); @@ -1156,24 +1156,24 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, /* need source address above miyazawa*/ } - dst_hold(&rt->u.dst); - inet->cork.dst = &rt->u.dst; + dst_hold(&rt->dst); + inet->cork.dst = &rt->dst; inet->cork.fl = *fl; np->cork.hop_limit = hlimit; np->cork.tclass = tclass; mtu = np->pmtudisc == IPV6_PMTUDISC_PROBE ? - rt->u.dst.dev->mtu : dst_mtu(rt->u.dst.path); + rt->dst.dev->mtu : dst_mtu(rt->dst.path); if (np->frag_size < mtu) { if (np->frag_size) mtu = np->frag_size; } inet->cork.fragsize = mtu; - if (dst_allfrag(rt->u.dst.path)) + if (dst_allfrag(rt->dst.path)) inet->cork.flags |= IPCORK_ALLFRAG; inet->cork.length = 0; sk->sk_sndmsg_page = NULL; sk->sk_sndmsg_off = 0; - exthdrlen = rt->u.dst.header_len + (opt ? opt->opt_flen : 0) - + exthdrlen = rt->dst.header_len + (opt ? opt->opt_flen : 0) - rt->rt6i_nfheader_len; length += exthdrlen; transhdrlen += exthdrlen; @@ -1186,7 +1186,7 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, mtu = inet->cork.fragsize; } - hh_len = LL_RESERVED_SPACE(rt->u.dst.dev); + hh_len = LL_RESERVED_SPACE(rt->dst.dev); fragheaderlen = sizeof(struct ipv6hdr) + rt->rt6i_nfheader_len + (opt ? opt->opt_nflen : 0); @@ -1224,7 +1224,7 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, } if (proto == IPPROTO_UDP && - (rt->u.dst.dev->features & NETIF_F_UFO)) { + (rt->dst.dev->features & NETIF_F_UFO)) { err = ip6_ufo_append_data(sk, getfrag, from, length, hh_len, fragheaderlen, @@ -1270,7 +1270,7 @@ alloc_new_skb: fraglen = datalen + fragheaderlen; if ((flags & MSG_MORE) && - !(rt->u.dst.dev->features&NETIF_F_SG)) + !(rt->dst.dev->features&NETIF_F_SG)) alloclen = mtu; else alloclen = datalen + fragheaderlen; @@ -1281,7 +1281,7 @@ alloc_new_skb: * because we have no idea if we're the last one. */ if (datalen == length + fraggap) - alloclen += rt->u.dst.trailer_len; + alloclen += rt->dst.trailer_len; /* * We just reserve space for fragment header. @@ -1358,7 +1358,7 @@ alloc_new_skb: if (copy > length) copy = length; - if (!(rt->u.dst.dev->features&NETIF_F_SG)) { + if (!(rt->dst.dev->features&NETIF_F_SG)) { unsigned int off; off = skb->len; @@ -1503,7 +1503,7 @@ int ip6_push_pending_frames(struct sock *sk) skb->priority = sk->sk_priority; skb->mark = sk->sk_mark; - skb_dst_set(skb, dst_clone(&rt->u.dst)); + skb_dst_set(skb, dst_clone(&rt->dst)); IP6_UPD_PO_STATS(net, rt->rt6i_idev, IPSTATS_MIB_OUT, skb->len); if (proto == IPPROTO_ICMPV6) { struct inet6_dev *idev = ip6_dst_idev(skb_dst(skb)); diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 8f39893..0fd027f 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -552,7 +552,7 @@ ip4ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, if (ip_route_output_key(dev_net(skb->dev), &rt, &fl)) goto out; - skb2->dev = rt->u.dst.dev; + skb2->dev = rt->dst.dev; /* route "incoming" packet */ if (rt->rt_flags & RTCF_LOCAL) { @@ -562,7 +562,7 @@ ip4ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, fl.fl4_src = eiph->saddr; fl.fl4_tos = eiph->tos; if (ip_route_output_key(dev_net(skb->dev), &rt, &fl) || - rt->u.dst.dev->type != ARPHRD_TUNNEL) { + rt->dst.dev->type != ARPHRD_TUNNEL) { ip_rt_put(rt); goto out; } @@ -626,7 +626,7 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, icmpv6_send(skb2, rel_type, rel_code, rel_info); if (rt) - dst_release(&rt->u.dst); + dst_release(&rt->dst); kfree_skb(skb2); } @@ -1135,7 +1135,7 @@ static void ip6_tnl_link_config(struct ip6_tnl *t) if (dev->mtu < IPV6_MIN_MTU) dev->mtu = IPV6_MIN_MTU; } - dst_release(&rt->u.dst); + dst_release(&rt->dst); } } diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c index 3e36d15..d1444b9 100644 --- a/net/ipv6/mcast.c +++ b/net/ipv6/mcast.c @@ -158,7 +158,7 @@ int ipv6_sock_mc_join(struct sock *sk, int ifindex, const struct in6_addr *addr) rt = rt6_lookup(net, addr, NULL, 0, 0); if (rt) { dev = rt->rt6i_dev; - dst_release(&rt->u.dst); + dst_release(&rt->dst); } } else dev = dev_get_by_index_rcu(net, ifindex); @@ -248,7 +248,7 @@ static struct inet6_dev *ip6_mc_find_dev_rcu(struct net *net, if (rt) { dev = rt->rt6i_dev; dev_hold(dev); - dst_release(&rt->u.dst); + dst_release(&rt->dst); } } else dev = dev_get_by_index_rcu(net, ifindex); diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c index 0abdc24..1fc46fc 100644 --- a/net/ipv6/ndisc.c +++ b/net/ipv6/ndisc.c @@ -1229,7 +1229,7 @@ static void ndisc_router_discovery(struct sk_buff *skb) ND_PRINTK0(KERN_ERR "ICMPv6 RA: %s() got default router without neighbour.\n", __func__); - dst_release(&rt->u.dst); + dst_release(&rt->dst); in6_dev_put(in6_dev); return; } @@ -1244,7 +1244,7 @@ static void ndisc_router_discovery(struct sk_buff *skb) if (ra_msg->icmph.icmp6_hop_limit) { in6_dev->cnf.hop_limit = ra_msg->icmph.icmp6_hop_limit; if (rt) - rt->u.dst.metrics[RTAX_HOPLIMIT-1] = ra_msg->icmph.icmp6_hop_limit; + rt->dst.metrics[RTAX_HOPLIMIT-1] = ra_msg->icmph.icmp6_hop_limit; } skip_defrtr: @@ -1363,7 +1363,7 @@ skip_linkparms: in6_dev->cnf.mtu6 = mtu; if (rt) - rt->u.dst.metrics[RTAX_MTU-1] = mtu; + rt->dst.metrics[RTAX_MTU-1] = mtu; rt6_mtu_change(skb->dev, mtu); } @@ -1384,7 +1384,7 @@ skip_linkparms: } out: if (rt) - dst_release(&rt->u.dst); + dst_release(&rt->dst); else if (neigh) neigh_release(neigh); in6_dev_put(in6_dev); diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 968b964..e677937 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -611,23 +611,23 @@ static int rawv6_send_hdrinc(struct sock *sk, void *from, int length, int err; struct rt6_info *rt = (struct rt6_info *)*dstp; - if (length > rt->u.dst.dev->mtu) { - ipv6_local_error(sk, EMSGSIZE, fl, rt->u.dst.dev->mtu); + if (length > rt->dst.dev->mtu) { + ipv6_local_error(sk, EMSGSIZE, fl, rt->dst.dev->mtu); return -EMSGSIZE; } if (flags&MSG_PROBE) goto out; skb = sock_alloc_send_skb(sk, - length + LL_ALLOCATED_SPACE(rt->u.dst.dev) + 15, + length + LL_ALLOCATED_SPACE(rt->dst.dev) + 15, flags & MSG_DONTWAIT, &err); if (skb == NULL) goto error; - skb_reserve(skb, LL_RESERVED_SPACE(rt->u.dst.dev)); + skb_reserve(skb, LL_RESERVED_SPACE(rt->dst.dev)); skb->priority = sk->sk_priority; skb->mark = sk->sk_mark; - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); *dstp = NULL; skb_put(skb, length); @@ -643,7 +643,7 @@ static int rawv6_send_hdrinc(struct sock *sk, void *from, int length, IP6_UPD_PO_STATS(sock_net(sk), rt->rt6i_idev, IPSTATS_MIB_OUT, skb->len); err = NF_HOOK(NFPROTO_IPV6, NF_INET_LOCAL_OUT, skb, NULL, - rt->u.dst.dev, dst_output); + rt->dst.dev, dst_output); if (err > 0) err = net_xmit_errno(err); if (err) diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 252d761..f770285 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -126,16 +126,14 @@ static struct dst_ops ip6_dst_blackhole_ops = { }; static struct rt6_info ip6_null_entry_template = { - .u = { - .dst = { - .__refcnt = ATOMIC_INIT(1), - .__use = 1, - .obsolete = -1, - .error = -ENETUNREACH, - .metrics = { [RTAX_HOPLIMIT - 1] = 255, }, - .input = ip6_pkt_discard, - .output = ip6_pkt_discard_out, - } + .dst = { + .__refcnt = ATOMIC_INIT(1), + .__use = 1, + .obsolete = -1, + .error = -ENETUNREACH, + .metrics = { [RTAX_HOPLIMIT - 1] = 255, }, + .input = ip6_pkt_discard, + .output = ip6_pkt_discard_out, }, .rt6i_flags = (RTF_REJECT | RTF_NONEXTHOP), .rt6i_protocol = RTPROT_KERNEL, @@ -149,16 +147,14 @@ static int ip6_pkt_prohibit(struct sk_buff *skb); static int ip6_pkt_prohibit_out(struct sk_buff *skb); static struct rt6_info ip6_prohibit_entry_template = { - .u = { - .dst = { - .__refcnt = ATOMIC_INIT(1), - .__use = 1, - .obsolete = -1, - .error = -EACCES, - .metrics = { [RTAX_HOPLIMIT - 1] = 255, }, - .input = ip6_pkt_prohibit, - .output = ip6_pkt_prohibit_out, - } + .dst = { + .__refcnt = ATOMIC_INIT(1), + .__use = 1, + .obsolete = -1, + .error = -EACCES, + .metrics = { [RTAX_HOPLIMIT - 1] = 255, }, + .input = ip6_pkt_prohibit, + .output = ip6_pkt_prohibit_out, }, .rt6i_flags = (RTF_REJECT | RTF_NONEXTHOP), .rt6i_protocol = RTPROT_KERNEL, @@ -167,16 +163,14 @@ static struct rt6_info ip6_prohibit_entry_template = { }; static struct rt6_info ip6_blk_hole_entry_template = { - .u = { - .dst = { - .__refcnt = ATOMIC_INIT(1), - .__use = 1, - .obsolete = -1, - .error = -EINVAL, - .metrics = { [RTAX_HOPLIMIT - 1] = 255, }, - .input = dst_discard, - .output = dst_discard, - } + .dst = { + .__refcnt = ATOMIC_INIT(1), + .__use = 1, + .obsolete = -1, + .error = -EINVAL, + .metrics = { [RTAX_HOPLIMIT - 1] = 255, }, + .input = dst_discard, + .output = dst_discard, }, .rt6i_flags = (RTF_REJECT | RTF_NONEXTHOP), .rt6i_protocol = RTPROT_KERNEL, @@ -249,7 +243,7 @@ static inline struct rt6_info *rt6_device_match(struct net *net, if (!oif && ipv6_addr_any(saddr)) goto out; - for (sprt = rt; sprt; sprt = sprt->u.dst.rt6_next) { + for (sprt = rt; sprt; sprt = sprt->dst.rt6_next) { struct net_device *dev = sprt->rt6i_dev; if (oif) { @@ -407,10 +401,10 @@ static struct rt6_info *find_rr_leaf(struct fib6_node *fn, match = NULL; for (rt = rr_head; rt && rt->rt6i_metric == metric; - rt = rt->u.dst.rt6_next) + rt = rt->dst.rt6_next) match = find_match(rt, oif, strict, &mpri, match); for (rt = fn->leaf; rt && rt != rr_head && rt->rt6i_metric == metric; - rt = rt->u.dst.rt6_next) + rt = rt->dst.rt6_next) match = find_match(rt, oif, strict, &mpri, match); return match; @@ -432,7 +426,7 @@ static struct rt6_info *rt6_select(struct fib6_node *fn, int oif, int strict) if (!match && (strict & RT6_LOOKUP_F_REACHABLE)) { - struct rt6_info *next = rt0->u.dst.rt6_next; + struct rt6_info *next = rt0->dst.rt6_next; /* no entries matched; do round-robin */ if (!next || next->rt6i_metric != rt0->rt6i_metric) @@ -517,7 +511,7 @@ int rt6_route_rcv(struct net_device *dev, u8 *opt, int len, rt->rt6i_expires = jiffies + HZ * lifetime; rt->rt6i_flags |= RTF_EXPIRES; } - dst_release(&rt->u.dst); + dst_release(&rt->dst); } return 0; } @@ -555,7 +549,7 @@ restart: rt = rt6_device_match(net, rt, &fl->fl6_src, fl->oif, flags); BACKTRACK(net, &fl->fl6_src); out: - dst_use(&rt->u.dst, jiffies); + dst_use(&rt->dst, jiffies); read_unlock_bh(&table->tb6_lock); return rt; @@ -643,7 +637,7 @@ static struct rt6_info *rt6_alloc_cow(struct rt6_info *ort, struct in6_addr *dad ipv6_addr_copy(&rt->rt6i_dst.addr, daddr); rt->rt6i_dst.plen = 128; rt->rt6i_flags |= RTF_CACHE; - rt->u.dst.flags |= DST_HOST; + rt->dst.flags |= DST_HOST; #ifdef CONFIG_IPV6_SUBTREES if (rt->rt6i_src.plen && saddr) { @@ -677,7 +671,7 @@ static struct rt6_info *rt6_alloc_cow(struct rt6_info *ort, struct in6_addr *dad if (net_ratelimit()) printk(KERN_WARNING "Neighbour table overflow.\n"); - dst_free(&rt->u.dst); + dst_free(&rt->dst); return NULL; } rt->rt6i_nexthop = neigh; @@ -694,7 +688,7 @@ static struct rt6_info *rt6_alloc_clone(struct rt6_info *ort, struct in6_addr *d ipv6_addr_copy(&rt->rt6i_dst.addr, daddr); rt->rt6i_dst.plen = 128; rt->rt6i_flags |= RTF_CACHE; - rt->u.dst.flags |= DST_HOST; + rt->dst.flags |= DST_HOST; rt->rt6i_nexthop = neigh_clone(ort->rt6i_nexthop); } return rt; @@ -726,7 +720,7 @@ restart: rt->rt6i_flags & RTF_CACHE) goto out; - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); read_unlock_bh(&table->tb6_lock); if (!rt->rt6i_nexthop && !(rt->rt6i_flags & RTF_NONEXTHOP)) @@ -739,10 +733,10 @@ restart: #endif } - dst_release(&rt->u.dst); + dst_release(&rt->dst); rt = nrt ? : net->ipv6.ip6_null_entry; - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); if (nrt) { err = ip6_ins_rt(nrt); if (!err) @@ -756,7 +750,7 @@ restart: * Race condition! In the gap, when table->tb6_lock was * released someone could insert this route. Relookup. */ - dst_release(&rt->u.dst); + dst_release(&rt->dst); goto relookup; out: @@ -764,11 +758,11 @@ out: reachable = 0; goto restart_2; } - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); read_unlock_bh(&table->tb6_lock); out2: - rt->u.dst.lastuse = jiffies; - rt->u.dst.__use++; + rt->dst.lastuse = jiffies; + rt->dst.__use++; return rt; } @@ -835,15 +829,15 @@ int ip6_dst_blackhole(struct sock *sk, struct dst_entry **dstp, struct flowi *fl struct dst_entry *new = NULL; if (rt) { - new = &rt->u.dst; + new = &rt->dst; atomic_set(&new->__refcnt, 1); new->__use = 1; new->input = dst_discard; new->output = dst_discard; - memcpy(new->metrics, ort->u.dst.metrics, RTAX_MAX*sizeof(u32)); - new->dev = ort->u.dst.dev; + memcpy(new->metrics, ort->dst.metrics, RTAX_MAX*sizeof(u32)); + new->dev = ort->dst.dev; if (new->dev) dev_hold(new->dev); rt->rt6i_idev = ort->rt6i_idev; @@ -912,7 +906,7 @@ static void ip6_link_failure(struct sk_buff *skb) rt = (struct rt6_info *) skb_dst(skb); if (rt) { if (rt->rt6i_flags&RTF_CACHE) { - dst_set_expires(&rt->u.dst, 0); + dst_set_expires(&rt->dst, 0); rt->rt6i_flags |= RTF_EXPIRES; } else if (rt->rt6i_node && (rt->rt6i_flags & RTF_DEFAULT)) rt->rt6i_node->fn_sernum = -1; @@ -986,14 +980,14 @@ struct dst_entry *icmp6_dst_alloc(struct net_device *dev, rt->rt6i_dev = dev; rt->rt6i_idev = idev; rt->rt6i_nexthop = neigh; - atomic_set(&rt->u.dst.__refcnt, 1); - rt->u.dst.metrics[RTAX_HOPLIMIT-1] = 255; - rt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(rt->rt6i_dev); - rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->u.dst)); - rt->u.dst.output = ip6_output; + atomic_set(&rt->dst.__refcnt, 1); + rt->dst.metrics[RTAX_HOPLIMIT-1] = 255; + rt->dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(rt->rt6i_dev); + rt->dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->dst)); + rt->dst.output = ip6_output; #if 0 /* there's no chance to use these for ndisc */ - rt->u.dst.flags = ipv6_addr_type(addr) & IPV6_ADDR_UNICAST + rt->dst.flags = ipv6_addr_type(addr) & IPV6_ADDR_UNICAST ? DST_HOST : 0; ipv6_addr_copy(&rt->rt6i_dst.addr, addr); @@ -1001,14 +995,14 @@ struct dst_entry *icmp6_dst_alloc(struct net_device *dev, #endif spin_lock_bh(&icmp6_dst_lock); - rt->u.dst.next = icmp6_dst_gc_list; - icmp6_dst_gc_list = &rt->u.dst; + rt->dst.next = icmp6_dst_gc_list; + icmp6_dst_gc_list = &rt->dst; spin_unlock_bh(&icmp6_dst_lock); fib6_force_start_gc(net); out: - return &rt->u.dst; + return &rt->dst; } int icmp6_dst_gc(void) @@ -1159,7 +1153,7 @@ int ip6_route_add(struct fib6_config *cfg) goto out; } - rt->u.dst.obsolete = -1; + rt->dst.obsolete = -1; rt->rt6i_expires = (cfg->fc_flags & RTF_EXPIRES) ? jiffies + clock_t_to_jiffies(cfg->fc_expires) : 0; @@ -1171,16 +1165,16 @@ int ip6_route_add(struct fib6_config *cfg) addr_type = ipv6_addr_type(&cfg->fc_dst); if (addr_type & IPV6_ADDR_MULTICAST) - rt->u.dst.input = ip6_mc_input; + rt->dst.input = ip6_mc_input; else - rt->u.dst.input = ip6_forward; + rt->dst.input = ip6_forward; - rt->u.dst.output = ip6_output; + rt->dst.output = ip6_output; ipv6_addr_prefix(&rt->rt6i_dst.addr, &cfg->fc_dst, cfg->fc_dst_len); rt->rt6i_dst.plen = cfg->fc_dst_len; if (rt->rt6i_dst.plen == 128) - rt->u.dst.flags = DST_HOST; + rt->dst.flags = DST_HOST; #ifdef CONFIG_IPV6_SUBTREES ipv6_addr_prefix(&rt->rt6i_src.addr, &cfg->fc_src, cfg->fc_src_len); @@ -1208,9 +1202,9 @@ int ip6_route_add(struct fib6_config *cfg) goto out; } } - rt->u.dst.output = ip6_pkt_discard_out; - rt->u.dst.input = ip6_pkt_discard; - rt->u.dst.error = -ENETUNREACH; + rt->dst.output = ip6_pkt_discard_out; + rt->dst.input = ip6_pkt_discard; + rt->dst.error = -ENETUNREACH; rt->rt6i_flags = RTF_REJECT|RTF_NONEXTHOP; goto install_route; } @@ -1244,7 +1238,7 @@ int ip6_route_add(struct fib6_config *cfg) goto out; if (dev) { if (dev != grt->rt6i_dev) { - dst_release(&grt->u.dst); + dst_release(&grt->dst); goto out; } } else { @@ -1255,7 +1249,7 @@ int ip6_route_add(struct fib6_config *cfg) } if (!(grt->rt6i_flags&RTF_GATEWAY)) err = 0; - dst_release(&grt->u.dst); + dst_release(&grt->dst); if (err) goto out; @@ -1294,18 +1288,18 @@ install_route: goto out; } - rt->u.dst.metrics[type - 1] = nla_get_u32(nla); + rt->dst.metrics[type - 1] = nla_get_u32(nla); } } } - if (dst_metric(&rt->u.dst, RTAX_HOPLIMIT) == 0) - rt->u.dst.metrics[RTAX_HOPLIMIT-1] = -1; - if (!dst_mtu(&rt->u.dst)) - rt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(dev); - if (!dst_metric(&rt->u.dst, RTAX_ADVMSS)) - rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->u.dst)); - rt->u.dst.dev = dev; + if (dst_metric(&rt->dst, RTAX_HOPLIMIT) == 0) + rt->dst.metrics[RTAX_HOPLIMIT-1] = -1; + if (!dst_mtu(&rt->dst)) + rt->dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(dev); + if (!dst_metric(&rt->dst, RTAX_ADVMSS)) + rt->dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->dst)); + rt->dst.dev = dev; rt->rt6i_idev = idev; rt->rt6i_table = table; @@ -1319,7 +1313,7 @@ out: if (idev) in6_dev_put(idev); if (rt) - dst_free(&rt->u.dst); + dst_free(&rt->dst); return err; } @@ -1336,7 +1330,7 @@ static int __ip6_del_rt(struct rt6_info *rt, struct nl_info *info) write_lock_bh(&table->tb6_lock); err = fib6_del(rt, info); - dst_release(&rt->u.dst); + dst_release(&rt->dst); write_unlock_bh(&table->tb6_lock); @@ -1369,7 +1363,7 @@ static int ip6_route_del(struct fib6_config *cfg) &cfg->fc_src, cfg->fc_src_len); if (fn) { - for (rt = fn->leaf; rt; rt = rt->u.dst.rt6_next) { + for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) { if (cfg->fc_ifindex && (rt->rt6i_dev == NULL || rt->rt6i_dev->ifindex != cfg->fc_ifindex)) @@ -1379,7 +1373,7 @@ static int ip6_route_del(struct fib6_config *cfg) continue; if (cfg->fc_metric && cfg->fc_metric != rt->rt6i_metric) continue; - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); read_unlock_bh(&table->tb6_lock); return __ip6_del_rt(rt, &cfg->fc_nlinfo); @@ -1421,7 +1415,7 @@ static struct rt6_info *__ip6_route_redirect(struct net *net, read_lock_bh(&table->tb6_lock); fn = fib6_lookup(&table->tb6_root, &fl->fl6_dst, &fl->fl6_src); restart: - for (rt = fn->leaf; rt; rt = rt->u.dst.rt6_next) { + for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) { /* * Current route is on-link; redirect is always invalid. * @@ -1445,7 +1439,7 @@ restart: rt = net->ipv6.ip6_null_entry; BACKTRACK(net, &fl->fl6_src); out: - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); read_unlock_bh(&table->tb6_lock); @@ -1513,10 +1507,10 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *src, * Look, redirects are sent only in response to data packets, * so that this nexthop apparently is reachable. --ANK */ - dst_confirm(&rt->u.dst); + dst_confirm(&rt->dst); /* Duplicate redirect: silently ignore. */ - if (neigh == rt->u.dst.neighbour) + if (neigh == rt->dst.neighbour) goto out; nrt = ip6_rt_copy(rt); @@ -1529,20 +1523,20 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *src, ipv6_addr_copy(&nrt->rt6i_dst.addr, dest); nrt->rt6i_dst.plen = 128; - nrt->u.dst.flags |= DST_HOST; + nrt->dst.flags |= DST_HOST; ipv6_addr_copy(&nrt->rt6i_gateway, (struct in6_addr*)neigh->primary_key); nrt->rt6i_nexthop = neigh_clone(neigh); /* Reset pmtu, it may be better */ - nrt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(neigh->dev); - nrt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(dev_net(neigh->dev), - dst_mtu(&nrt->u.dst)); + nrt->dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(neigh->dev); + nrt->dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(dev_net(neigh->dev), + dst_mtu(&nrt->dst)); if (ip6_ins_rt(nrt)) goto out; - netevent.old = &rt->u.dst; - netevent.new = &nrt->u.dst; + netevent.old = &rt->dst; + netevent.new = &nrt->dst; call_netevent_notifiers(NETEVENT_REDIRECT, &netevent); if (rt->rt6i_flags&RTF_CACHE) { @@ -1551,7 +1545,7 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *src, } out: - dst_release(&rt->u.dst); + dst_release(&rt->dst); } /* @@ -1570,7 +1564,7 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr, if (rt == NULL) return; - if (pmtu >= dst_mtu(&rt->u.dst)) + if (pmtu >= dst_mtu(&rt->dst)) goto out; if (pmtu < IPV6_MIN_MTU) { @@ -1588,7 +1582,7 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr, They are sent only in response to data packets, so that this nexthop apparently is reachable. --ANK */ - dst_confirm(&rt->u.dst); + dst_confirm(&rt->dst); /* Host route. If it is static, it would be better not to override it, but add new one, so that @@ -1596,10 +1590,10 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr, would return automatically. */ if (rt->rt6i_flags & RTF_CACHE) { - rt->u.dst.metrics[RTAX_MTU-1] = pmtu; + rt->dst.metrics[RTAX_MTU-1] = pmtu; if (allfrag) - rt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG; - dst_set_expires(&rt->u.dst, net->ipv6.sysctl.ip6_rt_mtu_expires); + rt->dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG; + dst_set_expires(&rt->dst, net->ipv6.sysctl.ip6_rt_mtu_expires); rt->rt6i_flags |= RTF_MODIFIED|RTF_EXPIRES; goto out; } @@ -1615,9 +1609,9 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr, nrt = rt6_alloc_clone(rt, daddr); if (nrt) { - nrt->u.dst.metrics[RTAX_MTU-1] = pmtu; + nrt->dst.metrics[RTAX_MTU-1] = pmtu; if (allfrag) - nrt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG; + nrt->dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG; /* According to RFC 1981, detecting PMTU increase shouldn't be * happened within 5 mins, the recommended timer is 10 mins. @@ -1625,13 +1619,13 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr, * which is 10 mins. After 10 mins the decreased pmtu is expired * and detecting PMTU increase will be automatically happened. */ - dst_set_expires(&nrt->u.dst, net->ipv6.sysctl.ip6_rt_mtu_expires); + dst_set_expires(&nrt->dst, net->ipv6.sysctl.ip6_rt_mtu_expires); nrt->rt6i_flags |= RTF_DYNAMIC|RTF_EXPIRES; ip6_ins_rt(nrt); } out: - dst_release(&rt->u.dst); + dst_release(&rt->dst); } /* @@ -1644,18 +1638,18 @@ static struct rt6_info * ip6_rt_copy(struct rt6_info *ort) struct rt6_info *rt = ip6_dst_alloc(&net->ipv6.ip6_dst_ops); if (rt) { - rt->u.dst.input = ort->u.dst.input; - rt->u.dst.output = ort->u.dst.output; - - memcpy(rt->u.dst.metrics, ort->u.dst.metrics, RTAX_MAX*sizeof(u32)); - rt->u.dst.error = ort->u.dst.error; - rt->u.dst.dev = ort->u.dst.dev; - if (rt->u.dst.dev) - dev_hold(rt->u.dst.dev); + rt->dst.input = ort->dst.input; + rt->dst.output = ort->dst.output; + + memcpy(rt->dst.metrics, ort->dst.metrics, RTAX_MAX*sizeof(u32)); + rt->dst.error = ort->dst.error; + rt->dst.dev = ort->dst.dev; + if (rt->dst.dev) + dev_hold(rt->dst.dev); rt->rt6i_idev = ort->rt6i_idev; if (rt->rt6i_idev) in6_dev_hold(rt->rt6i_idev); - rt->u.dst.lastuse = jiffies; + rt->dst.lastuse = jiffies; rt->rt6i_expires = 0; ipv6_addr_copy(&rt->rt6i_gateway, &ort->rt6i_gateway); @@ -1689,14 +1683,14 @@ static struct rt6_info *rt6_get_route_info(struct net *net, if (!fn) goto out; - for (rt = fn->leaf; rt; rt = rt->u.dst.rt6_next) { + for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) { if (rt->rt6i_dev->ifindex != ifindex) continue; if ((rt->rt6i_flags & (RTF_ROUTEINFO|RTF_GATEWAY)) != (RTF_ROUTEINFO|RTF_GATEWAY)) continue; if (!ipv6_addr_equal(&rt->rt6i_gateway, gwaddr)) continue; - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); break; } out: @@ -1744,14 +1738,14 @@ struct rt6_info *rt6_get_dflt_router(struct in6_addr *addr, struct net_device *d return NULL; write_lock_bh(&table->tb6_lock); - for (rt = table->tb6_root.leaf; rt; rt=rt->u.dst.rt6_next) { + for (rt = table->tb6_root.leaf; rt; rt=rt->dst.rt6_next) { if (dev == rt->rt6i_dev && ((rt->rt6i_flags & (RTF_ADDRCONF | RTF_DEFAULT)) == (RTF_ADDRCONF | RTF_DEFAULT)) && ipv6_addr_equal(&rt->rt6i_gateway, addr)) break; } if (rt) - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); write_unlock_bh(&table->tb6_lock); return rt; } @@ -1790,9 +1784,9 @@ void rt6_purge_dflt_routers(struct net *net) restart: read_lock_bh(&table->tb6_lock); - for (rt = table->tb6_root.leaf; rt; rt = rt->u.dst.rt6_next) { + for (rt = table->tb6_root.leaf; rt; rt = rt->dst.rt6_next) { if (rt->rt6i_flags & (RTF_DEFAULT | RTF_ADDRCONF)) { - dst_hold(&rt->u.dst); + dst_hold(&rt->dst); read_unlock_bh(&table->tb6_lock); ip6_del_rt(rt); goto restart; @@ -1930,15 +1924,15 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev, dev_hold(net->loopback_dev); in6_dev_hold(idev); - rt->u.dst.flags = DST_HOST; - rt->u.dst.input = ip6_input; - rt->u.dst.output = ip6_output; + rt->dst.flags = DST_HOST; + rt->dst.input = ip6_input; + rt->dst.output = ip6_output; rt->rt6i_dev = net->loopback_dev; rt->rt6i_idev = idev; - rt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(rt->rt6i_dev); - rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->u.dst)); - rt->u.dst.metrics[RTAX_HOPLIMIT-1] = -1; - rt->u.dst.obsolete = -1; + rt->dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(rt->rt6i_dev); + rt->dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->dst)); + rt->dst.metrics[RTAX_HOPLIMIT-1] = -1; + rt->dst.obsolete = -1; rt->rt6i_flags = RTF_UP | RTF_NONEXTHOP; if (anycast) @@ -1947,7 +1941,7 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev, rt->rt6i_flags |= RTF_LOCAL; neigh = ndisc_get_neigh(rt->rt6i_dev, &rt->rt6i_gateway); if (IS_ERR(neigh)) { - dst_free(&rt->u.dst); + dst_free(&rt->dst); /* We are casting this because that is the return * value type. But an errno encoded pointer is the @@ -1962,7 +1956,7 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev, rt->rt6i_dst.plen = 128; rt->rt6i_table = fib6_get_table(net, RT6_TABLE_LOCAL); - atomic_set(&rt->u.dst.__refcnt, 1); + atomic_set(&rt->dst.__refcnt, 1); return rt; } @@ -2033,12 +2027,12 @@ static int rt6_mtu_change_route(struct rt6_info *rt, void *p_arg) PMTU discouvery. */ if (rt->rt6i_dev == arg->dev && - !dst_metric_locked(&rt->u.dst, RTAX_MTU) && - (dst_mtu(&rt->u.dst) >= arg->mtu || - (dst_mtu(&rt->u.dst) < arg->mtu && - dst_mtu(&rt->u.dst) == idev->cnf.mtu6))) { - rt->u.dst.metrics[RTAX_MTU-1] = arg->mtu; - rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, arg->mtu); + !dst_metric_locked(&rt->dst, RTAX_MTU) && + (dst_mtu(&rt->dst) >= arg->mtu || + (dst_mtu(&rt->dst) < arg->mtu && + dst_mtu(&rt->dst) == idev->cnf.mtu6))) { + rt->dst.metrics[RTAX_MTU-1] = arg->mtu; + rt->dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, arg->mtu); } return 0; } @@ -2252,20 +2246,20 @@ static int rt6_fill_node(struct net *net, #endif NLA_PUT_U32(skb, RTA_IIF, iif); } else if (dst) { - struct inet6_dev *idev = ip6_dst_idev(&rt->u.dst); + struct inet6_dev *idev = ip6_dst_idev(&rt->dst); struct in6_addr saddr_buf; if (ipv6_dev_get_saddr(net, idev ? idev->dev : NULL, dst, 0, &saddr_buf) == 0) NLA_PUT(skb, RTA_PREFSRC, 16, &saddr_buf); } - if (rtnetlink_put_metrics(skb, rt->u.dst.metrics) < 0) + if (rtnetlink_put_metrics(skb, rt->dst.metrics) < 0) goto nla_put_failure; - if (rt->u.dst.neighbour) - NLA_PUT(skb, RTA_GATEWAY, 16, &rt->u.dst.neighbour->primary_key); + if (rt->dst.neighbour) + NLA_PUT(skb, RTA_GATEWAY, 16, &rt->dst.neighbour->primary_key); - if (rt->u.dst.dev) + if (rt->dst.dev) NLA_PUT_U32(skb, RTA_OIF, rt->rt6i_dev->ifindex); NLA_PUT_U32(skb, RTA_PRIORITY, rt->rt6i_metric); @@ -2277,8 +2271,8 @@ static int rt6_fill_node(struct net *net, else expires = INT_MAX; - if (rtnl_put_cacheinfo(skb, &rt->u.dst, 0, 0, 0, - expires, rt->u.dst.error) < 0) + if (rtnl_put_cacheinfo(skb, &rt->dst, 0, 0, 0, + expires, rt->dst.error) < 0) goto nla_put_failure; return nlmsg_end(skb, nlh); @@ -2364,7 +2358,7 @@ static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void skb_reserve(skb, MAX_HEADER + sizeof(struct ipv6hdr)); rt = (struct rt6_info*) ip6_route_output(net, NULL, &fl); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); err = rt6_fill_node(net, skb, rt, &fl.fl6_dst, &fl.fl6_src, iif, RTM_NEWROUTE, NETLINK_CB(in_skb).pid, @@ -2416,12 +2410,12 @@ static int ip6_route_dev_notify(struct notifier_block *this, struct net *net = dev_net(dev); if (event == NETDEV_REGISTER && (dev->flags & IFF_LOOPBACK)) { - net->ipv6.ip6_null_entry->u.dst.dev = dev; + net->ipv6.ip6_null_entry->dst.dev = dev; net->ipv6.ip6_null_entry->rt6i_idev = in6_dev_get(dev); #ifdef CONFIG_IPV6_MULTIPLE_TABLES - net->ipv6.ip6_prohibit_entry->u.dst.dev = dev; + net->ipv6.ip6_prohibit_entry->dst.dev = dev; net->ipv6.ip6_prohibit_entry->rt6i_idev = in6_dev_get(dev); - net->ipv6.ip6_blk_hole_entry->u.dst.dev = dev; + net->ipv6.ip6_blk_hole_entry->dst.dev = dev; net->ipv6.ip6_blk_hole_entry->rt6i_idev = in6_dev_get(dev); #endif } @@ -2464,8 +2458,8 @@ static int rt6_info_route(struct rt6_info *rt, void *p_arg) seq_puts(m, "00000000000000000000000000000000"); } seq_printf(m, " %08x %08x %08x %08x %8s\n", - rt->rt6i_metric, atomic_read(&rt->u.dst.__refcnt), - rt->u.dst.__use, rt->rt6i_flags, + rt->rt6i_metric, atomic_read(&rt->dst.__refcnt), + rt->dst.__use, rt->rt6i_flags, rt->rt6i_dev ? rt->rt6i_dev->name : ""); return 0; } @@ -2646,9 +2640,9 @@ static int __net_init ip6_route_net_init(struct net *net) GFP_KERNEL); if (!net->ipv6.ip6_null_entry) goto out_ip6_dst_ops; - net->ipv6.ip6_null_entry->u.dst.path = + net->ipv6.ip6_null_entry->dst.path = (struct dst_entry *)net->ipv6.ip6_null_entry; - net->ipv6.ip6_null_entry->u.dst.ops = &net->ipv6.ip6_dst_ops; + net->ipv6.ip6_null_entry->dst.ops = &net->ipv6.ip6_dst_ops; #ifdef CONFIG_IPV6_MULTIPLE_TABLES net->ipv6.ip6_prohibit_entry = kmemdup(&ip6_prohibit_entry_template, @@ -2656,18 +2650,18 @@ static int __net_init ip6_route_net_init(struct net *net) GFP_KERNEL); if (!net->ipv6.ip6_prohibit_entry) goto out_ip6_null_entry; - net->ipv6.ip6_prohibit_entry->u.dst.path = + net->ipv6.ip6_prohibit_entry->dst.path = (struct dst_entry *)net->ipv6.ip6_prohibit_entry; - net->ipv6.ip6_prohibit_entry->u.dst.ops = &net->ipv6.ip6_dst_ops; + net->ipv6.ip6_prohibit_entry->dst.ops = &net->ipv6.ip6_dst_ops; net->ipv6.ip6_blk_hole_entry = kmemdup(&ip6_blk_hole_entry_template, sizeof(*net->ipv6.ip6_blk_hole_entry), GFP_KERNEL); if (!net->ipv6.ip6_blk_hole_entry) goto out_ip6_prohibit_entry; - net->ipv6.ip6_blk_hole_entry->u.dst.path = + net->ipv6.ip6_blk_hole_entry->dst.path = (struct dst_entry *)net->ipv6.ip6_blk_hole_entry; - net->ipv6.ip6_blk_hole_entry->u.dst.ops = &net->ipv6.ip6_dst_ops; + net->ipv6.ip6_blk_hole_entry->dst.ops = &net->ipv6.ip6_dst_ops; #endif net->ipv6.sysctl.flush_delay = 0; @@ -2742,12 +2736,12 @@ int __init ip6_route_init(void) /* Registering of the loopback is done before this portion of code, * the loopback reference in rt6_info will not be taken, do it * manually for init_net */ - init_net.ipv6.ip6_null_entry->u.dst.dev = init_net.loopback_dev; + init_net.ipv6.ip6_null_entry->dst.dev = init_net.loopback_dev; init_net.ipv6.ip6_null_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev); #ifdef CONFIG_IPV6_MULTIPLE_TABLES - init_net.ipv6.ip6_prohibit_entry->u.dst.dev = init_net.loopback_dev; + init_net.ipv6.ip6_prohibit_entry->dst.dev = init_net.loopback_dev; init_net.ipv6.ip6_prohibit_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev); - init_net.ipv6.ip6_blk_hole_entry->u.dst.dev = init_net.loopback_dev; + init_net.ipv6.ip6_blk_hole_entry->dst.dev = init_net.loopback_dev; init_net.ipv6.ip6_blk_hole_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev); #endif ret = fib6_init(); diff --git a/net/ipv6/sit.c b/net/ipv6/sit.c index 702c532..4699cd3 100644 --- a/net/ipv6/sit.c +++ b/net/ipv6/sit.c @@ -712,7 +712,7 @@ static netdev_tx_t ipip6_tunnel_xmit(struct sk_buff *skb, stats->tx_carrier_errors++; goto tx_error_icmp; } - tdev = rt->u.dst.dev; + tdev = rt->dst.dev; if (tdev == dev) { ip_rt_put(rt); @@ -721,7 +721,7 @@ static netdev_tx_t ipip6_tunnel_xmit(struct sk_buff *skb, } if (df) { - mtu = dst_mtu(&rt->u.dst) - sizeof(struct iphdr); + mtu = dst_mtu(&rt->dst) - sizeof(struct iphdr); if (mtu < 68) { stats->collisions++; @@ -780,7 +780,7 @@ static netdev_tx_t ipip6_tunnel_xmit(struct sk_buff *skb, memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt)); IPCB(skb)->flags = 0; skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* * Push down and install the IPIP header. @@ -829,7 +829,7 @@ static void ipip6_tunnel_bind_dev(struct net_device *dev) .proto = IPPROTO_IPV6 }; struct rtable *rt; if (!ip_route_output_key(dev_net(dev), &rt, &fl)) { - tdev = rt->u.dst.dev; + tdev = rt->dst.dev; ip_rt_put(rt); } dev->flags |= IFF_POINTOPOINT; diff --git a/net/l2tp/l2tp_ip.c b/net/l2tp/l2tp_ip.c index 0852512..226a0ae 100644 --- a/net/l2tp/l2tp_ip.c +++ b/net/l2tp/l2tp_ip.c @@ -348,7 +348,7 @@ static int l2tp_ip_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len sk->sk_state = TCP_ESTABLISHED; inet->inet_id = jiffies; - sk_dst_set(sk, &rt->u.dst); + sk_dst_set(sk, &rt->dst); write_lock_bh(&l2tp_ip_lock); hlist_del_init(&sk->sk_bind_node); @@ -496,9 +496,9 @@ static int l2tp_ip_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *m if (ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 0)) goto no_route; } - sk_setup_caps(sk, &rt->u.dst); + sk_setup_caps(sk, &rt->dst); } - skb_dst_set(skb, dst_clone(&rt->u.dst)); + skb_dst_set(skb, dst_clone(&rt->dst)); /* Queue the packet to IP for output */ rc = ip_queue_xmit(skb); diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c index 93c15a1..02b078e 100644 --- a/net/netfilter/ipvs/ip_vs_xmit.c +++ b/net/netfilter/ipvs/ip_vs_xmit.c @@ -90,10 +90,10 @@ __ip_vs_get_out_rt(struct ip_vs_conn *cp, u32 rtos) &dest->addr.ip); return NULL; } - __ip_vs_dst_set(dest, rtos, dst_clone(&rt->u.dst)); + __ip_vs_dst_set(dest, rtos, dst_clone(&rt->dst)); IP_VS_DBG(10, "new dst %pI4, refcnt=%d, rtos=%X\n", &dest->addr.ip, - atomic_read(&rt->u.dst.__refcnt), rtos); + atomic_read(&rt->dst.__refcnt), rtos); } spin_unlock(&dest->dst_lock); } else { @@ -148,10 +148,10 @@ __ip_vs_get_out_rt_v6(struct ip_vs_conn *cp) &dest->addr.in6); return NULL; } - __ip_vs_dst_set(dest, 0, dst_clone(&rt->u.dst)); + __ip_vs_dst_set(dest, 0, dst_clone(&rt->dst)); IP_VS_DBG(10, "new dst %pI6, refcnt=%d\n", &dest->addr.in6, - atomic_read(&rt->u.dst.__refcnt)); + atomic_read(&rt->dst.__refcnt)); } spin_unlock(&dest->dst_lock); } else { @@ -198,7 +198,7 @@ do { \ (skb)->ipvs_property = 1; \ skb_forward_csum(skb); \ NF_HOOK(pf, NF_INET_LOCAL_OUT, (skb), NULL, \ - (rt)->u.dst.dev, dst_output); \ + (rt)->dst.dev, dst_output); \ } while (0) @@ -245,7 +245,7 @@ ip_vs_bypass_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, } /* MTU checking */ - mtu = dst_mtu(&rt->u.dst); + mtu = dst_mtu(&rt->dst); if ((skb->len > mtu) && (iph->frag_off & htons(IP_DF))) { ip_rt_put(rt); icmp_send(skb, ICMP_DEST_UNREACH,ICMP_FRAG_NEEDED, htonl(mtu)); @@ -265,7 +265,7 @@ ip_vs_bypass_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, /* drop old route */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* Another hack: avoid icmp_send in ip_fragment */ skb->local_df = 1; @@ -309,9 +309,9 @@ ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, } /* MTU checking */ - mtu = dst_mtu(&rt->u.dst); + mtu = dst_mtu(&rt->dst); if (skb->len > mtu) { - dst_release(&rt->u.dst); + dst_release(&rt->dst); icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu); IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; @@ -323,13 +323,13 @@ ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, */ skb = skb_share_check(skb, GFP_ATOMIC); if (unlikely(skb == NULL)) { - dst_release(&rt->u.dst); + dst_release(&rt->dst); return NF_STOLEN; } /* drop old route */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* Another hack: avoid icmp_send in ip_fragment */ skb->local_df = 1; @@ -376,7 +376,7 @@ ip_vs_nat_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, goto tx_error_icmp; /* MTU checking */ - mtu = dst_mtu(&rt->u.dst); + mtu = dst_mtu(&rt->dst); if ((skb->len > mtu) && (iph->frag_off & htons(IP_DF))) { ip_rt_put(rt); icmp_send(skb, ICMP_DEST_UNREACH,ICMP_FRAG_NEEDED, htonl(mtu)); @@ -388,12 +388,12 @@ ip_vs_nat_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, if (!skb_make_writable(skb, sizeof(struct iphdr))) goto tx_error_put; - if (skb_cow(skb, rt->u.dst.dev->hard_header_len)) + if (skb_cow(skb, rt->dst.dev->hard_header_len)) goto tx_error_put; /* drop old route */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* mangle the packet */ if (pp->dnat_handler && !pp->dnat_handler(skb, pp, cp)) @@ -452,9 +452,9 @@ ip_vs_nat_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, goto tx_error_icmp; /* MTU checking */ - mtu = dst_mtu(&rt->u.dst); + mtu = dst_mtu(&rt->dst); if (skb->len > mtu) { - dst_release(&rt->u.dst); + dst_release(&rt->dst); icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu); IP_VS_DBG_RL_PKT(0, pp, skb, 0, "ip_vs_nat_xmit_v6(): frag needed for"); @@ -465,12 +465,12 @@ ip_vs_nat_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (!skb_make_writable(skb, sizeof(struct ipv6hdr))) goto tx_error_put; - if (skb_cow(skb, rt->u.dst.dev->hard_header_len)) + if (skb_cow(skb, rt->dst.dev->hard_header_len)) goto tx_error_put; /* drop old route */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* mangle the packet */ if (pp->dnat_handler && !pp->dnat_handler(skb, pp, cp)) @@ -498,7 +498,7 @@ tx_error: kfree_skb(skb); return NF_STOLEN; tx_error_put: - dst_release(&rt->u.dst); + dst_release(&rt->dst); goto tx_error; } #endif @@ -549,9 +549,9 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, if (!(rt = __ip_vs_get_out_rt(cp, RT_TOS(tos)))) goto tx_error_icmp; - tdev = rt->u.dst.dev; + tdev = rt->dst.dev; - mtu = dst_mtu(&rt->u.dst) - sizeof(struct iphdr); + mtu = dst_mtu(&rt->dst) - sizeof(struct iphdr); if (mtu < 68) { ip_rt_put(rt); IP_VS_DBG_RL("%s(): mtu less than 68\n", __func__); @@ -601,7 +601,7 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, /* drop old route */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* * Push down and install the IPIP header. @@ -615,7 +615,7 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, iph->daddr = rt->rt_dst; iph->saddr = rt->rt_src; iph->ttl = old_iph->ttl; - ip_select_ident(iph, &rt->u.dst, NULL); + ip_select_ident(iph, &rt->dst, NULL); /* Another hack: avoid icmp_send in ip_fragment */ skb->local_df = 1; @@ -660,12 +660,12 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (!rt) goto tx_error_icmp; - tdev = rt->u.dst.dev; + tdev = rt->dst.dev; - mtu = dst_mtu(&rt->u.dst) - sizeof(struct ipv6hdr); + mtu = dst_mtu(&rt->dst) - sizeof(struct ipv6hdr); /* TODO IPv6: do we need this check in IPv6? */ if (mtu < 1280) { - dst_release(&rt->u.dst); + dst_release(&rt->dst); IP_VS_DBG_RL("%s(): mtu less than 1280\n", __func__); goto tx_error; } @@ -674,7 +674,7 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (mtu < ntohs(old_iph->payload_len) + sizeof(struct ipv6hdr)) { icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu); - dst_release(&rt->u.dst); + dst_release(&rt->dst); IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } @@ -689,7 +689,7 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, struct sk_buff *new_skb = skb_realloc_headroom(skb, max_headroom); if (!new_skb) { - dst_release(&rt->u.dst); + dst_release(&rt->dst); kfree_skb(skb); IP_VS_ERR_RL("%s(): no memory\n", __func__); return NF_STOLEN; @@ -707,7 +707,7 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, /* drop old route */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* * Push down and install the IPIP header. @@ -760,7 +760,7 @@ ip_vs_dr_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, goto tx_error_icmp; /* MTU checking */ - mtu = dst_mtu(&rt->u.dst); + mtu = dst_mtu(&rt->dst); if ((iph->frag_off & htons(IP_DF)) && skb->len > mtu) { icmp_send(skb, ICMP_DEST_UNREACH,ICMP_FRAG_NEEDED, htonl(mtu)); ip_rt_put(rt); @@ -780,7 +780,7 @@ ip_vs_dr_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, /* drop old route */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* Another hack: avoid icmp_send in ip_fragment */ skb->local_df = 1; @@ -813,10 +813,10 @@ ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, goto tx_error_icmp; /* MTU checking */ - mtu = dst_mtu(&rt->u.dst); + mtu = dst_mtu(&rt->dst); if (skb->len > mtu) { icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu); - dst_release(&rt->u.dst); + dst_release(&rt->dst); IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; } @@ -827,13 +827,13 @@ ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, */ skb = skb_share_check(skb, GFP_ATOMIC); if (unlikely(skb == NULL)) { - dst_release(&rt->u.dst); + dst_release(&rt->dst); return NF_STOLEN; } /* drop old route */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); /* Another hack: avoid icmp_send in ip_fragment */ skb->local_df = 1; @@ -888,7 +888,7 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, goto tx_error_icmp; /* MTU checking */ - mtu = dst_mtu(&rt->u.dst); + mtu = dst_mtu(&rt->dst); if ((skb->len > mtu) && (ip_hdr(skb)->frag_off & htons(IP_DF))) { ip_rt_put(rt); icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu)); @@ -900,12 +900,12 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, if (!skb_make_writable(skb, offset)) goto tx_error_put; - if (skb_cow(skb, rt->u.dst.dev->hard_header_len)) + if (skb_cow(skb, rt->dst.dev->hard_header_len)) goto tx_error_put; /* drop the old route when skb is not shared */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); ip_vs_nat_icmp(skb, pp, cp, 0); @@ -963,9 +963,9 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, goto tx_error_icmp; /* MTU checking */ - mtu = dst_mtu(&rt->u.dst); + mtu = dst_mtu(&rt->dst); if (skb->len > mtu) { - dst_release(&rt->u.dst); + dst_release(&rt->dst); icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu); IP_VS_DBG_RL("%s(): frag needed\n", __func__); goto tx_error; @@ -975,12 +975,12 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (!skb_make_writable(skb, offset)) goto tx_error_put; - if (skb_cow(skb, rt->u.dst.dev->hard_header_len)) + if (skb_cow(skb, rt->dst.dev->hard_header_len)) goto tx_error_put; /* drop the old route when skb is not shared */ skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); + skb_dst_set(skb, &rt->dst); ip_vs_nat_icmp_v6(skb, pp, cp, 0); @@ -1001,7 +1001,7 @@ out: LeaveFunction(10); return rc; tx_error_put: - dst_release(&rt->u.dst); + dst_release(&rt->dst); goto tx_error; } #endif diff --git a/net/netfilter/nf_conntrack_h323_main.c b/net/netfilter/nf_conntrack_h323_main.c index 6eaee7c..b969025 100644 --- a/net/netfilter/nf_conntrack_h323_main.c +++ b/net/netfilter/nf_conntrack_h323_main.c @@ -734,11 +734,11 @@ static int callforward_do_filter(const union nf_inet_addr *src, if (!afinfo->route((struct dst_entry **)&rt1, &fl1)) { if (!afinfo->route((struct dst_entry **)&rt2, &fl2)) { if (rt1->rt_gateway == rt2->rt_gateway && - rt1->u.dst.dev == rt2->u.dst.dev) + rt1->dst.dev == rt2->dst.dev) ret = 1; - dst_release(&rt2->u.dst); + dst_release(&rt2->dst); } - dst_release(&rt1->u.dst); + dst_release(&rt1->dst); } break; } @@ -753,11 +753,11 @@ static int callforward_do_filter(const union nf_inet_addr *src, if (!afinfo->route((struct dst_entry **)&rt2, &fl2)) { if (!memcmp(&rt1->rt6i_gateway, &rt2->rt6i_gateway, sizeof(rt1->rt6i_gateway)) && - rt1->u.dst.dev == rt2->u.dst.dev) + rt1->dst.dev == rt2->dst.dev) ret = 1; - dst_release(&rt2->u.dst); + dst_release(&rt2->dst); } - dst_release(&rt1->u.dst); + dst_release(&rt1->dst); } break; } diff --git a/net/netfilter/nf_conntrack_netbios_ns.c b/net/netfilter/nf_conntrack_netbios_ns.c index 497b222..aadde01 100644 --- a/net/netfilter/nf_conntrack_netbios_ns.c +++ b/net/netfilter/nf_conntrack_netbios_ns.c @@ -61,7 +61,7 @@ static int help(struct sk_buff *skb, unsigned int protoff, goto out; rcu_read_lock(); - in_dev = __in_dev_get_rcu(rt->u.dst.dev); + in_dev = __in_dev_get_rcu(rt->dst.dev); if (in_dev != NULL) { for_primary_ifa(in_dev) { if (ifa->ifa_broadcast == iph->daddr) { diff --git a/net/netfilter/xt_TCPMSS.c b/net/netfilter/xt_TCPMSS.c index 62ec021..1841388 100644 --- a/net/netfilter/xt_TCPMSS.c +++ b/net/netfilter/xt_TCPMSS.c @@ -165,8 +165,8 @@ static u_int32_t tcpmss_reverse_mtu(const struct sk_buff *skb, rcu_read_unlock(); if (rt != NULL) { - mtu = dst_mtu(&rt->u.dst); - dst_release(&rt->u.dst); + mtu = dst_mtu(&rt->dst); + dst_release(&rt->dst); } return mtu; } diff --git a/net/netfilter/xt_TEE.c b/net/netfilter/xt_TEE.c index 859d9fd..c77a85b 100644 --- a/net/netfilter/xt_TEE.c +++ b/net/netfilter/xt_TEE.c @@ -77,8 +77,8 @@ tee_tg_route4(struct sk_buff *skb, const struct xt_tee_tginfo *info) return false; skb_dst_drop(skb); - skb_dst_set(skb, &rt->u.dst); - skb->dev = rt->u.dst.dev; + skb_dst_set(skb, &rt->dst); + skb->dev = rt->dst.dev; skb->protocol = htons(ETH_P_IP); return true; } diff --git a/net/rxrpc/ar-peer.c b/net/rxrpc/ar-peer.c index f0f85b0..9f1729b 100644 --- a/net/rxrpc/ar-peer.c +++ b/net/rxrpc/ar-peer.c @@ -64,8 +64,8 @@ static void rxrpc_assess_MTU_size(struct rxrpc_peer *peer) return; } - peer->if_mtu = dst_mtu(&rt->u.dst); - dst_release(&rt->u.dst); + peer->if_mtu = dst_mtu(&rt->dst); + dst_release(&rt->dst); _leave(" [if_mtu %u]", peer->if_mtu); } diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index 1827498..a0e1a7f 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -490,7 +490,7 @@ static struct dst_entry *sctp_v4_get_dst(struct sctp_association *asoc, __func__, &fl.fl4_dst, &fl.fl4_src); if (!ip_route_output_key(&init_net, &rt, &fl)) { - dst = &rt->u.dst; + dst = &rt->dst; } /* If there is no association or if a source address is passed, no @@ -534,7 +534,7 @@ static struct dst_entry *sctp_v4_get_dst(struct sctp_association *asoc, fl.fl4_src = laddr->a.v4.sin_addr.s_addr; fl.fl_ip_sport = laddr->a.v4.sin_port; if (!ip_route_output_key(&init_net, &rt, &fl)) { - dst = &rt->u.dst; + dst = &rt->dst; goto out_unlock; } } -- cgit v1.1 From c7de2cf053420d63bac85133469c965d4b1083e1 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 9 Jun 2010 02:09:23 +0000 Subject: pkt_sched: gen_kill_estimator() rcu fixes gen_kill_estimator() API is incomplete or not well documented, since caller should make sure an RCU grace period is respected before freeing stats_lock. This was partially addressed in commit 5d944c640b4 (gen_estimator: deadlock fix), but same problem exist for all gen_kill_estimator() users, if lock they use is not already RCU protected. A code review shows xt_RATEEST.c, act_api.c, act_police.c have this problem. Other are ok because they use qdisc lock, already RCU protected. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/gen_estimator.c | 1 + net/netfilter/xt_RATEEST.c | 12 +++++++++++- net/sched/act_api.c | 11 ++++++++++- net/sched/act_police.c | 12 +++++++++++- 4 files changed, 33 insertions(+), 3 deletions(-) (limited to 'net') diff --git a/net/core/gen_estimator.c b/net/core/gen_estimator.c index 785e527..9fbe7f7 100644 --- a/net/core/gen_estimator.c +++ b/net/core/gen_estimator.c @@ -263,6 +263,7 @@ static void __gen_kill_estimator(struct rcu_head *head) * * Removes the rate estimator specified by &bstats and &rate_est. * + * Note : Caller should respect an RCU grace period before freeing stats_lock */ void gen_kill_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_rate_est *rate_est) diff --git a/net/netfilter/xt_RATEEST.c b/net/netfilter/xt_RATEEST.c index 69c01e1..de079abd 100644 --- a/net/netfilter/xt_RATEEST.c +++ b/net/netfilter/xt_RATEEST.c @@ -60,13 +60,22 @@ struct xt_rateest *xt_rateest_lookup(const char *name) } EXPORT_SYMBOL_GPL(xt_rateest_lookup); +static void xt_rateest_free_rcu(struct rcu_head *head) +{ + kfree(container_of(head, struct xt_rateest, rcu)); +} + void xt_rateest_put(struct xt_rateest *est) { mutex_lock(&xt_rateest_mutex); if (--est->refcnt == 0) { hlist_del(&est->list); gen_kill_estimator(&est->bstats, &est->rstats); - kfree(est); + /* + * gen_estimator est_timer() might access est->lock or bstats, + * wait a RCU grace period before freeing 'est' + */ + call_rcu(&est->rcu, xt_rateest_free_rcu); } mutex_unlock(&xt_rateest_mutex); } @@ -179,6 +188,7 @@ static int __init xt_rateest_tg_init(void) static void __exit xt_rateest_tg_fini(void) { xt_unregister_target(&xt_rateest_tg_reg); + rcu_barrier(); /* Wait for completion of call_rcu()'s (xt_rateest_free_rcu) */ } diff --git a/net/sched/act_api.c b/net/sched/act_api.c index 972378f..23b25f8 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -26,6 +26,11 @@ #include #include +static void tcf_common_free_rcu(struct rcu_head *head) +{ + kfree(container_of(head, struct tcf_common, tcfc_rcu)); +} + void tcf_hash_destroy(struct tcf_common *p, struct tcf_hashinfo *hinfo) { unsigned int h = tcf_hash(p->tcfc_index, hinfo->hmask); @@ -38,7 +43,11 @@ void tcf_hash_destroy(struct tcf_common *p, struct tcf_hashinfo *hinfo) write_unlock_bh(hinfo->lock); gen_kill_estimator(&p->tcfc_bstats, &p->tcfc_rate_est); - kfree(p); + /* + * gen_estimator est_timer() might access p->tcfc_lock + * or bstats, wait a RCU grace period before freeing p + */ + call_rcu(&p->tcfc_rcu, tcf_common_free_rcu); return; } } diff --git a/net/sched/act_police.c b/net/sched/act_police.c index 654f73d..537a487 100644 --- a/net/sched/act_police.c +++ b/net/sched/act_police.c @@ -97,6 +97,11 @@ nla_put_failure: goto done; } +static void tcf_police_free_rcu(struct rcu_head *head) +{ + kfree(container_of(head, struct tcf_police, tcf_rcu)); +} + static void tcf_police_destroy(struct tcf_police *p) { unsigned int h = tcf_hash(p->tcf_index, POL_TAB_MASK); @@ -113,7 +118,11 @@ static void tcf_police_destroy(struct tcf_police *p) qdisc_put_rtab(p->tcfp_R_tab); if (p->tcfp_P_tab) qdisc_put_rtab(p->tcfp_P_tab); - kfree(p); + /* + * gen_estimator est_timer() might access p->tcf_lock + * or bstats, wait a RCU grace period before freeing p + */ + call_rcu(&p->tcf_rcu, tcf_police_free_rcu); return; } } @@ -397,6 +406,7 @@ static void __exit police_cleanup_module(void) { tcf_unregister_action(&act_police_ops); + rcu_barrier(); /* Wait for completion of call_rcu()'s (tcf_police_free_rcu) */ } module_init(police_init_module); -- cgit v1.1 From 0c78a92fbd655ab990e2799f645707f05f548e2f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 9 Jun 2010 16:33:05 +0000 Subject: econet: fix locking econet lacks proper locking. It holds econet_lock only when inserting or deleting an entry in econet_sklist, not during lookups. - convert econet_lock from rwlock to spinlock - use econet_lock in ec_listening_socket() lookup - use appropriate sock_hold() / sock_put() to avoid corruptions. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/econet/af_econet.c | 27 ++++++++++++++++++--------- 1 file changed, 18 insertions(+), 9 deletions(-) (limited to 'net') diff --git a/net/econet/af_econet.c b/net/econet/af_econet.c index 2a5a805..dc54bd0 100644 --- a/net/econet/af_econet.c +++ b/net/econet/af_econet.c @@ -48,7 +48,7 @@ static const struct proto_ops econet_ops; static struct hlist_head econet_sklist; -static DEFINE_RWLOCK(econet_lock); +static DEFINE_SPINLOCK(econet_lock); static DEFINE_MUTEX(econet_mutex); /* Since there are only 256 possible network numbers (or fewer, depends @@ -98,16 +98,16 @@ struct ec_cb static void econet_remove_socket(struct hlist_head *list, struct sock *sk) { - write_lock_bh(&econet_lock); + spin_lock_bh(&econet_lock); sk_del_node_init(sk); - write_unlock_bh(&econet_lock); + spin_unlock_bh(&econet_lock); } static void econet_insert_socket(struct hlist_head *list, struct sock *sk) { - write_lock_bh(&econet_lock); + spin_lock_bh(&econet_lock); sk_add_node(sk, list); - write_unlock_bh(&econet_lock); + spin_unlock_bh(&econet_lock); } /* @@ -782,15 +782,19 @@ static struct sock *ec_listening_socket(unsigned char port, unsigned char struct sock *sk; struct hlist_node *node; + spin_lock(&econet_lock); sk_for_each(sk, node, &econet_sklist) { struct econet_sock *opt = ec_sk(sk); if ((opt->port == port || opt->port == 0) && (opt->station == station || opt->station == 0) && - (opt->net == net || opt->net == 0)) + (opt->net == net || opt->net == 0)) { + sock_hold(sk); goto found; + } } sk = NULL; found: + spin_unlock(&econet_lock); return sk; } @@ -852,7 +856,7 @@ static void aun_incoming(struct sk_buff *skb, struct aunhdr *ah, size_t len) { struct iphdr *ip = ip_hdr(skb); unsigned char stn = ntohl(ip->saddr) & 0xff; - struct sock *sk; + struct sock *sk = NULL; struct sk_buff *newskb; struct ec_device *edev = skb->dev->ec_ptr; @@ -882,10 +886,13 @@ static void aun_incoming(struct sk_buff *skb, struct aunhdr *ah, size_t len) } aun_send_response(ip->saddr, ah->handle, 3, 0); + sock_put(sk); return; bad: aun_send_response(ip->saddr, ah->handle, 4, 0); + if (sk) + sock_put(sk); } /* @@ -1050,7 +1057,7 @@ release: static int econet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct ec_framehdr *hdr; - struct sock *sk; + struct sock *sk = NULL; struct ec_device *edev = dev->ec_ptr; if (!net_eq(dev_net(dev), &init_net)) @@ -1085,10 +1092,12 @@ static int econet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet if (ec_queue_packet(sk, skb, edev->net, hdr->src_stn, hdr->cb, hdr->port)) goto drop; - + sock_put(sk); return NET_RX_SUCCESS; drop: + if (sk) + sock_put(sk); kfree_skb(skb); return NET_RX_DROP; } -- cgit v1.1 From 43d28b6515a6ea580a198df3e253e88236f08978 Mon Sep 17 00:00:00 2001 From: Daniel Turull Date: Wed, 9 Jun 2010 22:49:57 +0000 Subject: pktgen: increasing transmission granularity This patch increases the granularity of the rate generated by pktgen. The previous version of pktgen uses micro seconds (udelay) resolution when it was delayed causing gaps in the rates. It is changed to nanosecond (ndelay). Now any rate is possible. Also it allows to set, the desired rate in Mb/s or packets per second. The documentation has been updated. Signed-off-by: Daniel Turull Signed-off-by: Robert Olsson Signed-off-by: David S. Miller --- net/core/pktgen.c | 42 ++++++++++++++++++++++++++++++++++++++---- 1 file changed, 38 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/core/pktgen.c b/net/core/pktgen.c index 1dacd7b..6428653 100644 --- a/net/core/pktgen.c +++ b/net/core/pktgen.c @@ -169,7 +169,7 @@ #include #include /* do_div */ -#define VERSION "2.73" +#define VERSION "2.74" #define IP_NAME_SZ 32 #define MAX_MPLS_LABELS 16 /* This is the max label stack depth */ #define MPLS_STACK_BOTTOM htonl(0x00000100) @@ -980,6 +980,40 @@ static ssize_t pktgen_if_write(struct file *file, (unsigned long long) pkt_dev->delay); return count; } + if (!strcmp(name, "rate")) { + len = num_arg(&user_buffer[i], 10, &value); + if (len < 0) + return len; + + i += len; + if (!value) + return len; + pkt_dev->delay = pkt_dev->min_pkt_size*8*NSEC_PER_USEC/value; + if (debug) + printk(KERN_INFO + "pktgen: Delay set at: %llu ns\n", + pkt_dev->delay); + + sprintf(pg_result, "OK: rate=%lu", value); + return count; + } + if (!strcmp(name, "ratep")) { + len = num_arg(&user_buffer[i], 10, &value); + if (len < 0) + return len; + + i += len; + if (!value) + return len; + pkt_dev->delay = NSEC_PER_SEC/value; + if (debug) + printk(KERN_INFO + "pktgen: Delay set at: %llu ns\n", + pkt_dev->delay); + + sprintf(pg_result, "OK: rate=%lu", value); + return count; + } if (!strcmp(name, "udp_src_min")) { len = num_arg(&user_buffer[i], 10, &value); if (len < 0) @@ -2142,15 +2176,15 @@ static void spin(struct pktgen_dev *pkt_dev, ktime_t spin_until) hrtimer_init_on_stack(&t.timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); hrtimer_set_expires(&t.timer, spin_until); - remaining = ktime_to_us(hrtimer_expires_remaining(&t.timer)); + remaining = ktime_to_ns(hrtimer_expires_remaining(&t.timer)); if (remaining <= 0) { pkt_dev->next_tx = ktime_add_ns(spin_until, pkt_dev->delay); return; } start_time = ktime_now(); - if (remaining < 100) - udelay(remaining); /* really small just spin */ + if (remaining < 100000) + ndelay(remaining); /* really small just spin */ else { /* see do_nanosleep */ hrtimer_init_sleeper(&t, current); -- cgit v1.1 From be1f3c2c027cc5ad735df6a45a542ed1db7ec48b Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Tue, 8 Jun 2010 07:19:54 +0000 Subject: net: Enable 64-bit net device statistics on 32-bit architectures Use struct rtnl_link_stats64 as the statistics structure. On 32-bit architectures, insert 32 bits of padding after/before each field of struct net_device_stats to make its layout compatible with struct rtnl_link_stats64. Add an anonymous union in net_device; move stats into the union and add struct rtnl_link_stats64 stats64. Add net_device_ops::ndo_get_stats64, implementations of which will return a pointer to struct rtnl_link_stats64. Drivers that implement this operation must not update the structure asynchronously. Change dev_get_stats() to call ndo_get_stats64 if available, and to return a pointer to struct rtnl_link_stats64. Change callers of dev_get_stats() accordingly. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller --- net/8021q/vlanproc.c | 13 +++++++------ net/core/dev.c | 19 +++++++++++-------- net/core/net-sysfs.c | 12 ++++++------ net/core/rtnetlink.c | 6 +++--- 4 files changed, 27 insertions(+), 23 deletions(-) (limited to 'net') diff --git a/net/8021q/vlanproc.c b/net/8021q/vlanproc.c index afead35..df56f5c 100644 --- a/net/8021q/vlanproc.c +++ b/net/8021q/vlanproc.c @@ -278,8 +278,9 @@ static int vlandev_seq_show(struct seq_file *seq, void *offset) { struct net_device *vlandev = (struct net_device *) seq->private; const struct vlan_dev_info *dev_info = vlan_dev_info(vlandev); - const struct net_device_stats *stats; + const struct rtnl_link_stats64 *stats; static const char fmt[] = "%30s %12lu\n"; + static const char fmt64[] = "%30s %12llu\n"; int i; if (!is_vlan_dev(vlandev)) @@ -291,12 +292,12 @@ static int vlandev_seq_show(struct seq_file *seq, void *offset) vlandev->name, dev_info->vlan_id, (int)(dev_info->flags & 1), vlandev->priv_flags); - seq_printf(seq, fmt, "total frames received", stats->rx_packets); - seq_printf(seq, fmt, "total bytes received", stats->rx_bytes); - seq_printf(seq, fmt, "Broadcast/Multicast Rcvd", stats->multicast); + seq_printf(seq, fmt64, "total frames received", stats->rx_packets); + seq_printf(seq, fmt64, "total bytes received", stats->rx_bytes); + seq_printf(seq, fmt64, "Broadcast/Multicast Rcvd", stats->multicast); seq_puts(seq, "\n"); - seq_printf(seq, fmt, "total frames transmitted", stats->tx_packets); - seq_printf(seq, fmt, "total bytes transmitted", stats->tx_bytes); + seq_printf(seq, fmt64, "total frames transmitted", stats->tx_packets); + seq_printf(seq, fmt64, "total bytes transmitted", stats->tx_bytes); seq_printf(seq, fmt, "total headroom inc", dev_info->cnt_inc_headroom_on_tx); seq_printf(seq, fmt, "total encap on xmit", diff --git a/net/core/dev.c b/net/core/dev.c index 2778449..a1abc10 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3701,10 +3701,10 @@ void dev_seq_stop(struct seq_file *seq, void *v) static void dev_seq_printf_stats(struct seq_file *seq, struct net_device *dev) { - const struct net_device_stats *stats = dev_get_stats(dev); + const struct rtnl_link_stats64 *stats = dev_get_stats(dev); - seq_printf(seq, "%6s: %7lu %7lu %4lu %4lu %4lu %5lu %10lu %9lu " - "%8lu %7lu %4lu %4lu %4lu %5lu %7lu %10lu\n", + seq_printf(seq, "%6s: %7llu %7llu %4llu %4llu %4llu %5llu %10llu %9llu " + "%8llu %7llu %4llu %4llu %4llu %5llu %7llu %10llu\n", dev->name, stats->rx_bytes, stats->rx_packets, stats->rx_errors, stats->rx_dropped + stats->rx_missed_errors, @@ -5281,18 +5281,21 @@ EXPORT_SYMBOL(dev_txq_stats_fold); * @dev: device to get statistics from * * Get network statistics from device. The device driver may provide - * its own method by setting dev->netdev_ops->get_stats; otherwise - * the internal statistics structure is used. + * its own method by setting dev->netdev_ops->get_stats64 or + * dev->netdev_ops->get_stats; otherwise the internal statistics + * structure is used. */ -const struct net_device_stats *dev_get_stats(struct net_device *dev) +const struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev) { const struct net_device_ops *ops = dev->netdev_ops; + if (ops->ndo_get_stats64) + return ops->ndo_get_stats64(dev); if (ops->ndo_get_stats) - return ops->ndo_get_stats(dev); + return (struct rtnl_link_stats64 *)ops->ndo_get_stats(dev); dev_txq_stats_fold(dev, &dev->stats); - return &dev->stats; + return &dev->stats64; } EXPORT_SYMBOL(dev_get_stats); diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index 99e7052..ea3bb4c 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -29,6 +29,7 @@ static const char fmt_hex[] = "%#x\n"; static const char fmt_long_hex[] = "%#lx\n"; static const char fmt_dec[] = "%d\n"; static const char fmt_ulong[] = "%lu\n"; +static const char fmt_u64[] = "%llu\n"; static inline int dev_isalive(const struct net_device *dev) { @@ -324,14 +325,13 @@ static ssize_t netstat_show(const struct device *d, struct net_device *dev = to_net_dev(d); ssize_t ret = -EINVAL; - WARN_ON(offset > sizeof(struct net_device_stats) || - offset % sizeof(unsigned long) != 0); + WARN_ON(offset > sizeof(struct rtnl_link_stats64) || + offset % sizeof(u64) != 0); read_lock(&dev_base_lock); if (dev_isalive(dev)) { - const struct net_device_stats *stats = dev_get_stats(dev); - ret = sprintf(buf, fmt_ulong, - *(unsigned long *)(((u8 *) stats) + offset)); + const struct rtnl_link_stats64 *stats = dev_get_stats(dev); + ret = sprintf(buf, fmt_u64, *(u64 *)(((u8 *) stats) + offset)); } read_unlock(&dev_base_lock); return ret; @@ -343,7 +343,7 @@ static ssize_t show_##name(struct device *d, \ struct device_attribute *attr, char *buf) \ { \ return netstat_show(d, attr, buf, \ - offsetof(struct net_device_stats, name)); \ + offsetof(struct rtnl_link_stats64, name)); \ } \ static DEVICE_ATTR(name, S_IRUGO, show_##name, NULL) diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 1a2af24..e645778 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -579,7 +579,7 @@ static unsigned int rtnl_dev_combine_flags(const struct net_device *dev, } static void copy_rtnl_link_stats(struct rtnl_link_stats *a, - const struct net_device_stats *b) + const struct rtnl_link_stats64 *b) { a->rx_packets = b->rx_packets; a->tx_packets = b->tx_packets; @@ -610,7 +610,7 @@ static void copy_rtnl_link_stats(struct rtnl_link_stats *a, a->tx_compressed = b->tx_compressed; } -static void copy_rtnl_link_stats64(void *v, const struct net_device_stats *b) +static void copy_rtnl_link_stats64(void *v, const struct rtnl_link_stats64 *b) { struct rtnl_link_stats64 a; @@ -791,7 +791,7 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev, { struct ifinfomsg *ifm; struct nlmsghdr *nlh; - const struct net_device_stats *stats; + const struct rtnl_link_stats64 *stats; struct nlattr *attr; nlh = nlmsg_put(skb, pid, seq, type, sizeof(*ifm), flags); -- cgit v1.1 From 0b5c25e8ac3a60bd01a52ca7405ba96aec8c16be Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Tue, 8 Jun 2010 08:23:01 +0000 Subject: irttp: Print device parameters and statistics as unsigned Device statistics have type unsigned long and several of the device-specific parameters printed here have type __u32. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller --- net/irda/irttp.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'net') diff --git a/net/irda/irttp.c b/net/irda/irttp.c index 47db1d8..285761e 100644 --- a/net/irda/irttp.c +++ b/net/irda/irttp.c @@ -1853,23 +1853,23 @@ static int irttp_seq_show(struct seq_file *seq, void *v) self->remote_credit); seq_printf(seq, "send credit: %d\n", self->send_credit); - seq_printf(seq, " tx packets: %ld, ", + seq_printf(seq, " tx packets: %lu, ", self->stats.tx_packets); - seq_printf(seq, "rx packets: %ld, ", + seq_printf(seq, "rx packets: %lu, ", self->stats.rx_packets); - seq_printf(seq, "tx_queue len: %d ", + seq_printf(seq, "tx_queue len: %u ", skb_queue_len(&self->tx_queue)); - seq_printf(seq, "rx_queue len: %d\n", + seq_printf(seq, "rx_queue len: %u\n", skb_queue_len(&self->rx_queue)); seq_printf(seq, " tx_sdu_busy: %s, ", self->tx_sdu_busy? "TRUE":"FALSE"); seq_printf(seq, "rx_sdu_busy: %s\n", self->rx_sdu_busy? "TRUE":"FALSE"); - seq_printf(seq, " max_seg_size: %d, ", + seq_printf(seq, " max_seg_size: %u, ", self->max_seg_size); - seq_printf(seq, "tx_max_sdu_size: %d, ", + seq_printf(seq, "tx_max_sdu_size: %u, ", self->tx_max_sdu_size); - seq_printf(seq, "rx_max_sdu_size: %d\n", + seq_printf(seq, "rx_max_sdu_size: %u\n", self->rx_max_sdu_size); seq_printf(seq, " Used by (%s)\n\n", -- cgit v1.1 From f5c5440d40a24c5dc8030cde0a03debe87de4afb Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 14 Jun 2010 16:15:23 +0200 Subject: netfilter: nfnetlink_log: RCU conversion, part 2 - must use atomic_inc_not_zero() in instance_lookup_get() - must use hlist_add_head_rcu() instead of hlist_add_head() - must use hlist_del_rcu() instead of hlist_del() - Introduce NFULNL_COPY_DISABLED to stop lockless reader from using an instance, before we do final instance_put() on it. Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/netfilter/nfnetlink_log.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) (limited to 'net') diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c index 8ec23ec..fb86a51 100644 --- a/net/netfilter/nfnetlink_log.c +++ b/net/netfilter/nfnetlink_log.c @@ -109,8 +109,8 @@ instance_lookup_get(u_int16_t group_num) rcu_read_lock_bh(); inst = __instance_lookup(group_num); - if (inst) - instance_get(inst); + if (inst && !atomic_inc_not_zero(&inst->use)) + inst = NULL; rcu_read_unlock_bh(); return inst; @@ -171,7 +171,7 @@ instance_create(u_int16_t group_num, int pid) inst->copy_mode = NFULNL_COPY_PACKET; inst->copy_range = NFULNL_COPY_RANGE_MAX; - hlist_add_head(&inst->hlist, + hlist_add_head_rcu(&inst->hlist, &instance_table[instance_hashfn(group_num)]); spin_unlock_bh(&instances_lock); @@ -185,18 +185,23 @@ out_unlock: static void __nfulnl_flush(struct nfulnl_instance *inst); +/* called with BH disabled */ static void __instance_destroy(struct nfulnl_instance *inst) { /* first pull it out of the global list */ - hlist_del(&inst->hlist); + hlist_del_rcu(&inst->hlist); /* then flush all pending packets from skb */ - spin_lock_bh(&inst->lock); + spin_lock(&inst->lock); + + /* lockless readers wont be able to use us */ + inst->copy_mode = NFULNL_COPY_DISABLED; + if (inst->skb) __nfulnl_flush(inst); - spin_unlock_bh(&inst->lock); + spin_unlock(&inst->lock); /* and finally put the refcount */ instance_put(inst); @@ -624,6 +629,7 @@ nfulnl_log_packet(u_int8_t pf, size += nla_total_size(data_len); break; + case NFULNL_COPY_DISABLED: default: goto unlock_and_release; } -- cgit v1.1 From c86ee67c7c4bbab2818f653eb00a70671821624a Mon Sep 17 00:00:00 2001 From: Shan Wei Date: Mon, 14 Jun 2010 16:20:02 +0200 Subject: netfilter: kill redundant check code in which setting ip_summed value If the returned csum value is 0, We has set ip_summed with CHECKSUM_UNNECESSARY flag in __skb_checksum_complete_head(). So this patch kills the check and changes to return to upper caller directly. Signed-off-by: Shan Wei Signed-off-by: Patrick McHardy --- net/ipv4/netfilter.c | 4 +--- net/ipv6/netfilter.c | 4 +--- 2 files changed, 2 insertions(+), 6 deletions(-) (limited to 'net') diff --git a/net/ipv4/netfilter.c b/net/ipv4/netfilter.c index 07de855..acd1ea8 100644 --- a/net/ipv4/netfilter.c +++ b/net/ipv4/netfilter.c @@ -212,9 +212,7 @@ static __sum16 nf_ip_checksum_partial(struct sk_buff *skb, unsigned int hook, skb->csum = csum_tcpudp_nofold(iph->saddr, iph->daddr, protocol, skb->len - dataoff, 0); skb->ip_summed = CHECKSUM_NONE; - csum = __skb_checksum_complete_head(skb, dataoff + len); - if (!csum) - skb->ip_summed = CHECKSUM_UNNECESSARY; + return __skb_checksum_complete_head(skb, dataoff + len); } return csum; } diff --git a/net/ipv6/netfilter.c b/net/ipv6/netfilter.c index a74951c..7155b24 100644 --- a/net/ipv6/netfilter.c +++ b/net/ipv6/netfilter.c @@ -151,9 +151,7 @@ static __sum16 nf_ip6_checksum_partial(struct sk_buff *skb, unsigned int hook, protocol, csum_sub(0, hsum))); skb->ip_summed = CHECKSUM_NONE; - csum = __skb_checksum_complete_head(skb, dataoff + len); - if (!csum) - skb->ip_summed = CHECKSUM_UNNECESSARY; + return __skb_checksum_complete_head(skb, dataoff + len); } return csum; }; -- cgit v1.1 From 841a5940eb872d70dad2b9ee7f946d8fd13a8c22 Mon Sep 17 00:00:00 2001 From: Shan Wei Date: Mon, 14 Jun 2010 16:28:23 +0200 Subject: netfilter: defrag: remove one redundant atomic ops Instead of doing one atomic operation per frag, we can factorize them. Reported from Eric Dumazet. Signed-off-by: Shan Wei Acked-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/ipv6/netfilter/nf_conntrack_reasm.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'net') diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c index 6fb8901..bc5b86d 100644 --- a/net/ipv6/netfilter/nf_conntrack_reasm.c +++ b/net/ipv6/netfilter/nf_conntrack_reasm.c @@ -442,7 +442,6 @@ nf_ct_frag6_reasm(struct nf_ct_frag6_queue *fq, struct net_device *dev) skb_shinfo(head)->frag_list = head->next; skb_reset_transport_header(head); skb_push(head, head->data - skb_network_header(head)); - atomic_sub(head->truesize, &nf_init_frags.mem); for (fp=head->next; fp; fp = fp->next) { head->data_len += fp->len; @@ -452,8 +451,8 @@ nf_ct_frag6_reasm(struct nf_ct_frag6_queue *fq, struct net_device *dev) else if (head->ip_summed == CHECKSUM_COMPLETE) head->csum = csum_add(head->csum, fp->csum); head->truesize += fp->truesize; - atomic_sub(fp->truesize, &nf_init_frags.mem); } + atomic_sub(head->truesize, &nf_init_frags.mem); head->next = NULL; head->dev = dev; -- cgit v1.1 From 0b041f8d1e6fb11a6134d37230da8c2182f99110 Mon Sep 17 00:00:00 2001 From: Shan Wei Date: Mon, 14 Jun 2010 16:30:47 +0200 Subject: netfilter: defrag: kill unused work parameter of frag_kfree_skb() The parameter (work) is unused, remove it. Reported from Eric Dumazet. Signed-off-by: Shan Wei Acked-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/ipv6/netfilter/nf_conntrack_reasm.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c index bc5b86d..9254008 100644 --- a/net/ipv6/netfilter/nf_conntrack_reasm.c +++ b/net/ipv6/netfilter/nf_conntrack_reasm.c @@ -114,10 +114,8 @@ static void nf_skb_free(struct sk_buff *skb) } /* Memory Tracking Functions. */ -static inline void frag_kfree_skb(struct sk_buff *skb, unsigned int *work) +static void frag_kfree_skb(struct sk_buff *skb) { - if (work) - *work -= skb->truesize; atomic_sub(skb->truesize, &nf_init_frags.mem); nf_skb_free(skb); kfree_skb(skb); @@ -335,7 +333,7 @@ static int nf_ct_frag6_queue(struct nf_ct_frag6_queue *fq, struct sk_buff *skb, fq->q.fragments = next; fq->q.meat -= free_it->len; - frag_kfree_skb(free_it, NULL); + frag_kfree_skb(free_it); } } -- cgit v1.1 From 9d38d85de0270e3927bffab94973a9c78d1dc800 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 9 Jun 2010 17:20:33 +0200 Subject: cfg80211/mac80211: allow action frame TX/RX in IBSS When in IBSS mode, currently action frame TX and RX cannot be used. Allow using it to talk to any peer, or for public action frames. Also, while at it, restructure the code in mac80211 to make it easier to add this for other interface types in the future. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/cfg.c | 52 +++++++++++++++++++++++++++++++++++++++++++--- net/mac80211/ieee80211_i.h | 5 ----- net/mac80211/mlme.c | 38 --------------------------------- net/mac80211/rx.c | 3 +-- net/wireless/mlme.c | 5 +++-- net/wireless/nl80211.c | 6 ++++-- 6 files changed, 57 insertions(+), 52 deletions(-) (limited to 'net') diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 952845e..59f597d 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1555,9 +1555,55 @@ static int ieee80211_action(struct wiphy *wiphy, struct net_device *dev, bool channel_type_valid, const u8 *buf, size_t len, u64 *cookie) { - return ieee80211_mgd_action(IEEE80211_DEV_TO_SUB_IF(dev), chan, - channel_type, channel_type_valid, - buf, len, cookie); + struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); + struct ieee80211_local *local = sdata->local; + struct sk_buff *skb; + struct sta_info *sta; + const struct ieee80211_mgmt *mgmt = (void *)buf; + u32 flags = IEEE80211_TX_INTFL_NL80211_FRAME_TX | + IEEE80211_TX_CTL_REQ_TX_STATUS; + + /* Check that we are on the requested channel for transmission */ + if (chan != local->tmp_channel && + chan != local->oper_channel) + return -EBUSY; + if (channel_type_valid && + (channel_type != local->tmp_channel_type && + channel_type != local->_oper_channel_type)) + return -EBUSY; + + switch (sdata->vif.type) { + case NL80211_IFTYPE_ADHOC: + if (mgmt->u.action.category == WLAN_CATEGORY_PUBLIC) + break; + rcu_read_lock(); + sta = sta_info_get(sdata, mgmt->da); + rcu_read_unlock(); + if (!sta) + return -ENOLINK; + break; + case NL80211_IFTYPE_STATION: + if (!(sdata->u.mgd.flags & IEEE80211_STA_MFP_ENABLED)) + flags |= IEEE80211_TX_INTFL_DONT_ENCRYPT; + break; + default: + return -EOPNOTSUPP; + } + + skb = dev_alloc_skb(local->hw.extra_tx_headroom + len); + if (!skb) + return -ENOMEM; + skb_reserve(skb, local->hw.extra_tx_headroom); + + memcpy(skb_put(skb, len), buf, len); + + IEEE80211_SKB_CB(skb)->flags = flags; + + skb->dev = sdata->dev; + ieee80211_tx_skb(sdata, skb); + + *cookie = (unsigned long) skb; + return 0; } struct cfg80211_ops mac80211_config_ops = { diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 4d3883e..5782a53 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -986,11 +986,6 @@ int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, int ieee80211_mgd_disassoc(struct ieee80211_sub_if_data *sdata, struct cfg80211_disassoc_request *req, void *cookie); -int ieee80211_mgd_action(struct ieee80211_sub_if_data *sdata, - struct ieee80211_channel *chan, - enum nl80211_channel_type channel_type, - bool channel_type_valid, - const u8 *buf, size_t len, u64 *cookie); ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb); void ieee80211_send_pspoll(struct ieee80211_local *local, diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 2ab4e86..1373b3d 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2353,44 +2353,6 @@ int ieee80211_mgd_disassoc(struct ieee80211_sub_if_data *sdata, return 0; } -int ieee80211_mgd_action(struct ieee80211_sub_if_data *sdata, - struct ieee80211_channel *chan, - enum nl80211_channel_type channel_type, - bool channel_type_valid, - const u8 *buf, size_t len, u64 *cookie) -{ - struct ieee80211_local *local = sdata->local; - struct ieee80211_if_managed *ifmgd = &sdata->u.mgd; - struct sk_buff *skb; - - /* Check that we are on the requested channel for transmission */ - if (chan != local->tmp_channel && - chan != local->oper_channel) - return -EBUSY; - if (channel_type_valid && - (channel_type != local->tmp_channel_type && - channel_type != local->_oper_channel_type)) - return -EBUSY; - - skb = dev_alloc_skb(local->hw.extra_tx_headroom + len); - if (!skb) - return -ENOMEM; - skb_reserve(skb, local->hw.extra_tx_headroom); - - memcpy(skb_put(skb, len), buf, len); - - if (!(ifmgd->flags & IEEE80211_STA_MFP_ENABLED)) - IEEE80211_SKB_CB(skb)->flags |= - IEEE80211_TX_INTFL_DONT_ENCRYPT; - IEEE80211_SKB_CB(skb)->flags |= IEEE80211_TX_INTFL_NL80211_FRAME_TX | - IEEE80211_TX_CTL_REQ_TX_STATUS; - skb->dev = sdata->dev; - ieee80211_tx_skb(sdata, skb); - - *cookie = (unsigned long) skb; - return 0; -} - void ieee80211_cqm_rssi_notify(struct ieee80211_vif *vif, enum nl80211_cqm_rssi_threshold_event rssi_event, gfp_t gfp) diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 6e7d6d4..3adcda5 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -2045,8 +2045,7 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) */ status = IEEE80211_SKB_RXCB(rx->skb); - if (sdata->vif.type == NL80211_IFTYPE_STATION && - cfg80211_rx_action(rx->sdata->dev, status->freq, + if (cfg80211_rx_action(rx->sdata->dev, status->freq, rx->skb->data, rx->skb->len, GFP_ATOMIC)) goto handled; diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index f69ae19..9f95354 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -846,8 +846,9 @@ int cfg80211_mlme_action(struct cfg80211_registered_device *rdev, if (!wdev->current_bss || memcmp(wdev->current_bss->pub.bssid, mgmt->bssid, ETH_ALEN) != 0 || - memcmp(wdev->current_bss->pub.bssid, mgmt->da, - ETH_ALEN) != 0) + (wdev->iftype == NL80211_IFTYPE_STATION && + memcmp(wdev->current_bss->pub.bssid, mgmt->da, + ETH_ALEN) != 0)) return -ENOTCONN; } diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 90ab3c8..c65e67e 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -4653,7 +4653,8 @@ static int nl80211_register_action(struct sk_buff *skb, struct genl_info *info) if (err) goto unlock_rtnl; - if (dev->ieee80211_ptr->iftype != NL80211_IFTYPE_STATION) { + if (dev->ieee80211_ptr->iftype != NL80211_IFTYPE_STATION && + dev->ieee80211_ptr->iftype != NL80211_IFTYPE_ADHOC) { err = -EOPNOTSUPP; goto out; } @@ -4703,7 +4704,8 @@ static int nl80211_action(struct sk_buff *skb, struct genl_info *info) goto out; } - if (dev->ieee80211_ptr->iftype != NL80211_IFTYPE_STATION) { + if (dev->ieee80211_ptr->iftype != NL80211_IFTYPE_STATION && + dev->ieee80211_ptr->iftype != NL80211_IFTYPE_ADHOC) { err = -EOPNOTSUPP; goto out; } -- cgit v1.1 From 2a419056c15478d2df3f3e9d4fa64e34eb1faa7d Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:29 +0200 Subject: mac80211: simplify station/aggregation code A number of places use RCU locking for accessing the station list, even though they do not need to. Use mutex locking instead to prepare for the locking changes I want to make. The mlme code is also using a WLAN_STA_DISASSOC flag that has the same meaning as WLAN_STA_BLOCK_BA, so use that. While doing so, combine places where we loop over stations twice, and optimise away some of the loops by checking if the hardware supports aggregation at all first. Also fix a more theoretical race condition: right now we could resume, set up an aggregation session, and right after tear it down again due to the code that is needed for hardware reconfiguration here. Also mark add a comment to that code marking it as a workaround. Finally, remove a pointless aggregation disabling loop when an interface is stopped, directly after that we remove all stations from it which will also disable all aggregation sessions that may still be active, and does so in a race-free way unlike the current loop that doesn't block new sessions. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 10 +--------- net/mac80211/iface.c | 13 ------------- net/mac80211/mlme.c | 6 +++--- net/mac80211/pm.c | 16 ++++------------ net/mac80211/sta_info.h | 4 ---- net/mac80211/util.c | 31 ++++++++++++++++--------------- 6 files changed, 24 insertions(+), 56 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index d1b6664..9b9f21b 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -239,17 +239,9 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) sdata->vif.type != NL80211_IFTYPE_AP) return -EINVAL; - if (test_sta_flags(sta, WLAN_STA_DISASSOC)) { -#ifdef CONFIG_MAC80211_HT_DEBUG - printk(KERN_DEBUG "Disassociation is in progress. " - "Denying BA session request\n"); -#endif - return -EINVAL; - } - if (test_sta_flags(sta, WLAN_STA_BLOCK_BA)) { #ifdef CONFIG_MAC80211_HT_DEBUG - printk(KERN_DEBUG "Suspend in progress. " + printk(KERN_DEBUG "BA sessions blocked. " "Denying BA session request\n"); #endif return -EINVAL; diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 1afa9ec..906fc2b 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -339,7 +339,6 @@ static int ieee80211_stop(struct net_device *dev) { struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev); struct ieee80211_local *local = sdata->local; - struct sta_info *sta; unsigned long flags; struct sk_buff *skb, *tmp; u32 hw_reconf_flags = 0; @@ -356,18 +355,6 @@ static int ieee80211_stop(struct net_device *dev) ieee80211_work_purge(sdata); /* - * Now delete all active aggregation sessions. - */ - rcu_read_lock(); - - list_for_each_entry_rcu(sta, &local->sta_list, list) { - if (sta->sdata == sdata) - ieee80211_sta_tear_down_BA_sessions(sta); - } - - rcu_read_unlock(); - - /* * Remove all stations associated with this interface. * * This must be done before calling ops->remove_interface() diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 1373b3d..0154d74 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -898,13 +898,13 @@ static void ieee80211_set_disassoc(struct ieee80211_sub_if_data *sdata, netif_tx_stop_all_queues(sdata->dev); netif_carrier_off(sdata->dev); - rcu_read_lock(); + mutex_lock(&local->sta_mtx); sta = sta_info_get(sdata, bssid); if (sta) { - set_sta_flags(sta, WLAN_STA_DISASSOC); + set_sta_flags(sta, WLAN_STA_BLOCK_BA); ieee80211_sta_tear_down_BA_sessions(sta); } - rcu_read_unlock(); + mutex_unlock(&local->sta_mtx); changed |= ieee80211_reset_erp_info(sdata); diff --git a/net/mac80211/pm.c b/net/mac80211/pm.c index 75202b2..e145a94 100644 --- a/net/mac80211/pm.c +++ b/net/mac80211/pm.c @@ -40,22 +40,14 @@ int __ieee80211_suspend(struct ieee80211_hw *hw) list_for_each_entry(sdata, &local->interfaces, list) ieee80211_disable_keys(sdata); - /* Tear down aggregation sessions */ - - rcu_read_lock(); - - if (hw->flags & IEEE80211_HW_AMPDU_AGGREGATION) { - list_for_each_entry_rcu(sta, &local->sta_list, list) { + /* tear down aggregation sessions and remove STAs */ + mutex_lock(&local->sta_mtx); + list_for_each_entry(sta, &local->sta_list, list) { + if (hw->flags & IEEE80211_HW_AMPDU_AGGREGATION) { set_sta_flags(sta, WLAN_STA_BLOCK_BA); ieee80211_sta_tear_down_BA_sessions(sta); } - } - - rcu_read_unlock(); - /* remove STAs */ - mutex_lock(&local->sta_mtx); - list_for_each_entry(sta, &local->sta_list, list) { if (sta->uploaded) { sdata = sta->sdata; if (sdata->vif.type == NL80211_IFTYPE_AP_VLAN) diff --git a/net/mac80211/sta_info.h b/net/mac80211/sta_info.h index 813da34..786bbd3 100644 --- a/net/mac80211/sta_info.h +++ b/net/mac80211/sta_info.h @@ -42,9 +42,6 @@ * be in the queues * @WLAN_STA_PSPOLL: Station sent PS-poll while driver was keeping * station in power-save mode, reply when the driver unblocks. - * @WLAN_STA_DISASSOC: Disassociation in progress. - * This is used to reject TX BA session requests when disassociation - * is in progress. */ enum ieee80211_sta_info_flags { WLAN_STA_AUTH = 1<<0, @@ -60,7 +57,6 @@ enum ieee80211_sta_info_flags { WLAN_STA_BLOCK_BA = 1<<11, WLAN_STA_PS_DRIVER = 1<<12, WLAN_STA_PSPOLL = 1<<13, - WLAN_STA_DISASSOC = 1<<14, }; #define STA_TID_NUM 16 diff --git a/net/mac80211/util.c b/net/mac80211/util.c index 5b79d55..a54cf14 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -1138,18 +1138,6 @@ int ieee80211_reconfig(struct ieee80211_local *local) } mutex_unlock(&local->sta_mtx); - /* Clear Suspend state so that ADDBA requests can be processed */ - - rcu_read_lock(); - - if (hw->flags & IEEE80211_HW_AMPDU_AGGREGATION) { - list_for_each_entry_rcu(sta, &local->sta_list, list) { - clear_sta_flags(sta, WLAN_STA_BLOCK_BA); - } - } - - rcu_read_unlock(); - /* setup RTS threshold */ drv_set_rts_threshold(local, hw->wiphy->rts_threshold); @@ -1202,13 +1190,26 @@ int ieee80211_reconfig(struct ieee80211_local *local) } } - rcu_read_lock(); + /* + * Clear the WLAN_STA_BLOCK_BA flag so new aggregation + * sessions can be established after a resume. + * + * Also tear down aggregation sessions since reconfiguring + * them in a hardware restart scenario is not easily done + * right now, and the hardware will have lost information + * about the sessions, but we and the AP still think they + * are active. This is really a workaround though. + */ if (hw->flags & IEEE80211_HW_AMPDU_AGGREGATION) { - list_for_each_entry_rcu(sta, &local->sta_list, list) { + mutex_lock(&local->sta_mtx); + + list_for_each_entry(sta, &local->sta_list, list) { ieee80211_sta_tear_down_BA_sessions(sta); + clear_sta_flags(sta, WLAN_STA_BLOCK_BA); } + + mutex_unlock(&local->sta_mtx); } - rcu_read_unlock(); /* add back keys */ list_for_each_entry(sdata, &local->interfaces, list) -- cgit v1.1 From 35f20c14a103ca2c7062999e934a513b83d84de6 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:30 +0200 Subject: mac80211: use common skb queue IBSS, managed and mesh modes all have an skb queue, and in the future we want to also use it in other modes, so make them all use a common skb queue already. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/ibss.c | 7 +++---- net/mac80211/ieee80211_i.h | 7 ++----- net/mac80211/iface.c | 21 +++++++++++---------- net/mac80211/mesh.c | 14 ++------------ net/mac80211/mlme.c | 5 ++--- 5 files changed, 20 insertions(+), 34 deletions(-) (limited to 'net') diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index d7a96ce..a9ff904 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -779,7 +779,7 @@ static void ieee80211_ibss_work(struct work_struct *work) return; ifibss = &sdata->u.ibss; - while ((skb = skb_dequeue(&ifibss->skb_queue))) + while ((skb = skb_dequeue(&sdata->skb_queue))) ieee80211_ibss_rx_queued_mgmt(sdata, skb); if (!test_and_clear_bit(IEEE80211_IBSS_REQ_RUN, &ifibss->request)) @@ -850,7 +850,6 @@ void ieee80211_ibss_setup_sdata(struct ieee80211_sub_if_data *sdata) INIT_WORK(&ifibss->work, ieee80211_ibss_work); setup_timer(&ifibss->timer, ieee80211_ibss_timer, (unsigned long) sdata); - skb_queue_head_init(&ifibss->skb_queue); } /* scan finished notification */ @@ -890,7 +889,7 @@ ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) case IEEE80211_STYPE_BEACON: case IEEE80211_STYPE_PROBE_REQ: case IEEE80211_STYPE_AUTH: - skb_queue_tail(&sdata->u.ibss.skb_queue, skb); + skb_queue_tail(&sdata->skb_queue, skb); ieee80211_queue_work(&local->hw, &sdata->u.ibss.work); return RX_QUEUED; } @@ -983,7 +982,7 @@ int ieee80211_ibss_leave(struct ieee80211_sub_if_data *sdata) synchronize_rcu(); kfree_skb(skb); - skb_queue_purge(&sdata->u.ibss.skb_queue); + skb_queue_purge(&sdata->skb_queue); memset(sdata->u.ibss.bssid, 0, ETH_ALEN); sdata->u.ibss.ssid_len = 0; diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 5782a53..3cc3867 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -340,8 +340,6 @@ struct ieee80211_if_managed { u16 aid; - struct sk_buff_head skb_queue; - unsigned long timers_running; /* used for quiesce/restart */ bool powersave; /* powersave requested for this iface */ enum ieee80211_smps_mode req_smps, /* requested smps mode */ @@ -388,8 +386,6 @@ struct ieee80211_if_ibss { struct timer_list timer; struct work_struct work; - struct sk_buff_head skb_queue; - unsigned long request; unsigned long last_scan_completed; @@ -420,7 +416,6 @@ struct ieee80211_if_mesh { struct timer_list housekeeping_timer; struct timer_list mesh_path_timer; struct timer_list mesh_path_root_timer; - struct sk_buff_head skb_queue; unsigned long timers_running; @@ -517,6 +512,8 @@ struct ieee80211_sub_if_data { u16 sequence_number; + struct sk_buff_head skb_queue; + /* * AP this belongs to: self in AP mode and * corresponding AP in VLAN mode, NULL for diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 906fc2b..56167a3 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -465,21 +465,11 @@ static int ieee80211_stop(struct net_device *dev) cancel_work_sync(&sdata->u.mgd.monitor_work); cancel_work_sync(&sdata->u.mgd.beacon_connection_loss_work); - /* - * When we get here, the interface is marked down. - * Call synchronize_rcu() to wait for the RX path - * should it be using the interface and enqueuing - * frames at this very time on another CPU. - */ - synchronize_rcu(); - skb_queue_purge(&sdata->u.mgd.skb_queue); /* fall through */ case NL80211_IFTYPE_ADHOC: if (sdata->vif.type == NL80211_IFTYPE_ADHOC) { del_timer_sync(&sdata->u.ibss.timer); cancel_work_sync(&sdata->u.ibss.work); - synchronize_rcu(); - skb_queue_purge(&sdata->u.ibss.skb_queue); } /* fall through */ case NL80211_IFTYPE_MESH_POINT: @@ -495,6 +485,15 @@ static int ieee80211_stop(struct net_device *dev) } /* fall through */ default: + /* + * When we get here, the interface is marked down. + * Call synchronize_rcu() to wait for the RX path + * should it be using the interface and enqueuing + * frames at this very time on another CPU. + */ + synchronize_rcu(); + skb_queue_purge(&sdata->skb_queue); + if (local->scan_sdata == sdata) ieee80211_scan_cancel(local); @@ -721,6 +720,8 @@ static void ieee80211_setup_sdata(struct ieee80211_sub_if_data *sdata, /* only monitor differs */ sdata->dev->type = ARPHRD_ETHER; + skb_queue_head_init(&sdata->skb_queue); + switch (type) { case NL80211_IFTYPE_AP: skb_queue_head_init(&sdata->u.ap.ps_bc_buf); diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index bde8103..065533a 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -537,15 +537,6 @@ void ieee80211_stop_mesh(struct ieee80211_sub_if_data *sdata) * it no longer is. */ cancel_work_sync(&sdata->u.mesh.work); - - /* - * When we get here, the interface is marked down. - * Call synchronize_rcu() to wait for the RX path - * should it be using the interface and enqueuing - * frames at this very time on another CPU. - */ - rcu_barrier(); /* Wait for RX path and call_rcu()'s */ - skb_queue_purge(&sdata->u.mesh.skb_queue); } static void ieee80211_mesh_rx_bcn_presp(struct ieee80211_sub_if_data *sdata, @@ -650,7 +641,7 @@ static void ieee80211_mesh_work(struct work_struct *work) if (local->scanning) return; - while ((skb = skb_dequeue(&ifmsh->skb_queue))) + while ((skb = skb_dequeue(&sdata->skb_queue))) ieee80211_mesh_rx_queued_mgmt(sdata, skb); if (ifmsh->preq_queue_len && @@ -690,7 +681,6 @@ void ieee80211_mesh_init_sdata(struct ieee80211_sub_if_data *sdata) setup_timer(&ifmsh->housekeeping_timer, ieee80211_mesh_housekeeping_timer, (unsigned long) sdata); - skb_queue_head_init(&sdata->u.mesh.skb_queue); ifmsh->mshcfg.dot11MeshRetryTimeout = MESH_RET_T; ifmsh->mshcfg.dot11MeshConfirmTimeout = MESH_CONF_T; @@ -750,7 +740,7 @@ ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) case IEEE80211_STYPE_ACTION: case IEEE80211_STYPE_PROBE_RESP: case IEEE80211_STYPE_BEACON: - skb_queue_tail(&ifmsh->skb_queue, skb); + skb_queue_tail(&sdata->skb_queue, skb); ieee80211_queue_work(&local->hw, &ifmsh->work); return RX_QUEUED; } diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 0154d74..854ef4e 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1652,7 +1652,7 @@ ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, case IEEE80211_STYPE_DEAUTH: case IEEE80211_STYPE_DISASSOC: case IEEE80211_STYPE_ACTION: - skb_queue_tail(&sdata->u.mgd.skb_queue, skb); + skb_queue_tail(&sdata->skb_queue, skb); ieee80211_queue_work(&local->hw, &sdata->u.mgd.work); return RX_QUEUED; } @@ -1810,7 +1810,7 @@ static void ieee80211_sta_work(struct work_struct *work) ifmgd = &sdata->u.mgd; /* first process frames to avoid timing out while a frame is pending */ - while ((skb = skb_dequeue(&ifmgd->skb_queue))) + while ((skb = skb_dequeue(&sdata->skb_queue))) ieee80211_sta_rx_queued_mgmt(sdata, skb); /* then process the rest of the work */ @@ -1967,7 +1967,6 @@ void ieee80211_sta_setup_sdata(struct ieee80211_sub_if_data *sdata) (unsigned long) sdata); setup_timer(&ifmgd->chswitch_timer, ieee80211_chswitch_timer, (unsigned long) sdata); - skb_queue_head_init(&ifmgd->skb_queue); ifmgd->flags = 0; -- cgit v1.1 From 64592c8fc0e99d445fc3fdedddeb6088e20086f1 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:31 +0200 Subject: mac80211: use common work struct IBSS, managed and mesh modes all have their own work struct, and in the future we want to also use it in other modes to process frames from the now common skb queue. This also makes the skb queue and work safe to use from other interface types. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/ibss.c | 13 ++++++------- net/mac80211/ieee80211_i.h | 4 +--- net/mac80211/iface.c | 6 ++---- net/mac80211/mesh.c | 22 +++++++++------------- net/mac80211/mesh_hwmp.c | 4 ++-- net/mac80211/mesh_pathtbl.c | 4 ++-- net/mac80211/mlme.c | 12 +++++------- net/mac80211/pm.c | 2 ++ 8 files changed, 29 insertions(+), 38 deletions(-) (limited to 'net') diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index a9ff904..18c4266 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -761,7 +761,7 @@ static void ieee80211_ibss_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, static void ieee80211_ibss_work(struct work_struct *work) { struct ieee80211_sub_if_data *sdata = - container_of(work, struct ieee80211_sub_if_data, u.ibss.work); + container_of(work, struct ieee80211_sub_if_data, work); struct ieee80211_local *local = sdata->local; struct ieee80211_if_ibss *ifibss; struct sk_buff *skb; @@ -804,7 +804,7 @@ static void ieee80211_queue_ibss_work(struct ieee80211_sub_if_data *sdata) struct ieee80211_local *local = sdata->local; set_bit(IEEE80211_IBSS_REQ_RUN, &ifibss->request); - ieee80211_queue_work(&local->hw, &ifibss->work); + ieee80211_queue_work(&local->hw, &sdata->work); } static void ieee80211_ibss_timer(unsigned long data) @@ -827,7 +827,6 @@ void ieee80211_ibss_quiesce(struct ieee80211_sub_if_data *sdata) { struct ieee80211_if_ibss *ifibss = &sdata->u.ibss; - cancel_work_sync(&ifibss->work); if (del_timer_sync(&ifibss->timer)) ifibss->timer_running = true; } @@ -847,7 +846,7 @@ void ieee80211_ibss_setup_sdata(struct ieee80211_sub_if_data *sdata) { struct ieee80211_if_ibss *ifibss = &sdata->u.ibss; - INIT_WORK(&ifibss->work, ieee80211_ibss_work); + INIT_WORK(&sdata->work, ieee80211_ibss_work); setup_timer(&ifibss->timer, ieee80211_ibss_timer, (unsigned long) sdata); } @@ -890,7 +889,7 @@ ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) case IEEE80211_STYPE_PROBE_REQ: case IEEE80211_STYPE_AUTH: skb_queue_tail(&sdata->skb_queue, skb); - ieee80211_queue_work(&local->hw, &sdata->u.ibss.work); + ieee80211_queue_work(&local->hw, &sdata->work); return RX_QUEUED; } @@ -956,7 +955,7 @@ int ieee80211_ibss_join(struct ieee80211_sub_if_data *sdata, ieee80211_recalc_idle(sdata->local); set_bit(IEEE80211_IBSS_REQ_RUN, &sdata->u.ibss.request); - ieee80211_queue_work(&sdata->local->hw, &sdata->u.ibss.work); + ieee80211_queue_work(&sdata->local->hw, &sdata->work); return 0; } @@ -967,7 +966,7 @@ int ieee80211_ibss_leave(struct ieee80211_sub_if_data *sdata) del_timer_sync(&sdata->u.ibss.timer); clear_bit(IEEE80211_IBSS_REQ_RUN, &sdata->u.ibss.request); - cancel_work_sync(&sdata->u.ibss.work); + cancel_work_sync(&sdata->work); clear_bit(IEEE80211_IBSS_REQ_RUN, &sdata->u.ibss.request); sta_info_flush(sdata->local, sdata); diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 3cc3867..2873f63 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -325,7 +325,6 @@ struct ieee80211_if_managed { struct timer_list conn_mon_timer; struct timer_list bcn_mon_timer; struct timer_list chswitch_timer; - struct work_struct work; struct work_struct monitor_work; struct work_struct chswitch_work; struct work_struct beacon_connection_loss_work; @@ -384,7 +383,6 @@ enum ieee80211_ibss_request { struct ieee80211_if_ibss { struct timer_list timer; - struct work_struct work; unsigned long request; unsigned long last_scan_completed; @@ -412,7 +410,6 @@ struct ieee80211_if_ibss { }; struct ieee80211_if_mesh { - struct work_struct work; struct timer_list housekeeping_timer; struct timer_list mesh_path_timer; struct timer_list mesh_path_root_timer; @@ -512,6 +509,7 @@ struct ieee80211_sub_if_data { u16 sequence_number; + struct work_struct work; struct sk_buff_head skb_queue; /* diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 56167a3..de7ddc3 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -460,17 +460,14 @@ static int ieee80211_stop(struct net_device *dev) * whether the interface is running, which, at this point, * it no longer is. */ - cancel_work_sync(&sdata->u.mgd.work); cancel_work_sync(&sdata->u.mgd.chswitch_work); cancel_work_sync(&sdata->u.mgd.monitor_work); cancel_work_sync(&sdata->u.mgd.beacon_connection_loss_work); /* fall through */ case NL80211_IFTYPE_ADHOC: - if (sdata->vif.type == NL80211_IFTYPE_ADHOC) { + if (sdata->vif.type == NL80211_IFTYPE_ADHOC) del_timer_sync(&sdata->u.ibss.timer); - cancel_work_sync(&sdata->u.ibss.work); - } /* fall through */ case NL80211_IFTYPE_MESH_POINT: if (ieee80211_vif_is_mesh(&sdata->vif)) { @@ -485,6 +482,7 @@ static int ieee80211_stop(struct net_device *dev) } /* fall through */ default: + cancel_work_sync(&sdata->work); /* * When we get here, the interface is marked down. * Call synchronize_rcu() to wait for the RX path diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index 065533a..be9aa98 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -54,7 +54,7 @@ static void ieee80211_mesh_housekeeping_timer(unsigned long data) return; } - ieee80211_queue_work(&local->hw, &ifmsh->work); + ieee80211_queue_work(&local->hw, &sdata->work); } /** @@ -345,7 +345,7 @@ static void ieee80211_mesh_path_timer(unsigned long data) return; } - ieee80211_queue_work(&local->hw, &ifmsh->work); + ieee80211_queue_work(&local->hw, &sdata->work); } static void ieee80211_mesh_path_root_timer(unsigned long data) @@ -362,7 +362,7 @@ static void ieee80211_mesh_path_root_timer(unsigned long data) return; } - ieee80211_queue_work(&local->hw, &ifmsh->work); + ieee80211_queue_work(&local->hw, &sdata->work); } void ieee80211_mesh_root_setup(struct ieee80211_if_mesh *ifmsh) @@ -484,9 +484,6 @@ void ieee80211_mesh_quiesce(struct ieee80211_sub_if_data *sdata) { struct ieee80211_if_mesh *ifmsh = &sdata->u.mesh; - /* might restart the timer but that doesn't matter */ - cancel_work_sync(&ifmsh->work); - /* use atomic bitops in case both timers fire at the same time */ if (del_timer_sync(&ifmsh->housekeeping_timer)) @@ -518,7 +515,7 @@ void ieee80211_start_mesh(struct ieee80211_sub_if_data *sdata) set_bit(MESH_WORK_HOUSEKEEPING, &ifmsh->wrkq_flags); ieee80211_mesh_root_setup(ifmsh); - ieee80211_queue_work(&local->hw, &ifmsh->work); + ieee80211_queue_work(&local->hw, &sdata->work); sdata->vif.bss_conf.beacon_int = MESH_DEFAULT_BEACON_INTERVAL; ieee80211_bss_info_change_notify(sdata, BSS_CHANGED_BEACON | BSS_CHANGED_BEACON_ENABLED | @@ -536,7 +533,7 @@ void ieee80211_stop_mesh(struct ieee80211_sub_if_data *sdata) * whether the interface is running, which, at this point, * it no longer is. */ - cancel_work_sync(&sdata->u.mesh.work); + cancel_work_sync(&sdata->work); } static void ieee80211_mesh_rx_bcn_presp(struct ieee80211_sub_if_data *sdata, @@ -630,7 +627,7 @@ static void ieee80211_mesh_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, static void ieee80211_mesh_work(struct work_struct *work) { struct ieee80211_sub_if_data *sdata = - container_of(work, struct ieee80211_sub_if_data, u.mesh.work); + container_of(work, struct ieee80211_sub_if_data, work); struct ieee80211_local *local = sdata->local; struct ieee80211_if_mesh *ifmsh = &sdata->u.mesh; struct sk_buff *skb; @@ -669,7 +666,7 @@ void ieee80211_mesh_notify_scan_completed(struct ieee80211_local *local) rcu_read_lock(); list_for_each_entry_rcu(sdata, &local->interfaces, list) if (ieee80211_vif_is_mesh(&sdata->vif)) - ieee80211_queue_work(&local->hw, &sdata->u.mesh.work); + ieee80211_queue_work(&local->hw, &sdata->work); rcu_read_unlock(); } @@ -677,7 +674,7 @@ void ieee80211_mesh_init_sdata(struct ieee80211_sub_if_data *sdata) { struct ieee80211_if_mesh *ifmsh = &sdata->u.mesh; - INIT_WORK(&ifmsh->work, ieee80211_mesh_work); + INIT_WORK(&sdata->work, ieee80211_mesh_work); setup_timer(&ifmsh->housekeeping_timer, ieee80211_mesh_housekeeping_timer, (unsigned long) sdata); @@ -726,7 +723,6 @@ ieee80211_rx_result ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) { struct ieee80211_local *local = sdata->local; - struct ieee80211_if_mesh *ifmsh = &sdata->u.mesh; struct ieee80211_mgmt *mgmt; u16 fc; @@ -741,7 +737,7 @@ ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) case IEEE80211_STYPE_PROBE_RESP: case IEEE80211_STYPE_BEACON: skb_queue_tail(&sdata->skb_queue, skb); - ieee80211_queue_work(&local->hw, &ifmsh->work); + ieee80211_queue_work(&local->hw, &sdata->work); return RX_QUEUED; } diff --git a/net/mac80211/mesh_hwmp.c b/net/mac80211/mesh_hwmp.c index 0705018..829e08a 100644 --- a/net/mac80211/mesh_hwmp.c +++ b/net/mac80211/mesh_hwmp.c @@ -805,14 +805,14 @@ static void mesh_queue_preq(struct mesh_path *mpath, u8 flags) spin_unlock(&ifmsh->mesh_preq_queue_lock); if (time_after(jiffies, ifmsh->last_preq + min_preq_int_jiff(sdata))) - ieee80211_queue_work(&sdata->local->hw, &ifmsh->work); + ieee80211_queue_work(&sdata->local->hw, &sdata->work); else if (time_before(jiffies, ifmsh->last_preq)) { /* avoid long wait if did not send preqs for a long time * and jiffies wrapped around */ ifmsh->last_preq = jiffies - min_preq_int_jiff(sdata) - 1; - ieee80211_queue_work(&sdata->local->hw, &ifmsh->work); + ieee80211_queue_work(&sdata->local->hw, &sdata->work); } else mod_timer(&ifmsh->mesh_path_timer, ifmsh->last_preq + min_preq_int_jiff(sdata)); diff --git a/net/mac80211/mesh_pathtbl.c b/net/mac80211/mesh_pathtbl.c index 181ffd6..349e466 100644 --- a/net/mac80211/mesh_pathtbl.c +++ b/net/mac80211/mesh_pathtbl.c @@ -315,7 +315,7 @@ int mesh_path_add(u8 *dst, struct ieee80211_sub_if_data *sdata) read_unlock(&pathtbl_resize_lock); if (grow) { set_bit(MESH_WORK_GROW_MPATH_TABLE, &ifmsh->wrkq_flags); - ieee80211_queue_work(&local->hw, &ifmsh->work); + ieee80211_queue_work(&local->hw, &sdata->work); } return 0; @@ -425,7 +425,7 @@ int mpp_path_add(u8 *dst, u8 *mpp, struct ieee80211_sub_if_data *sdata) read_unlock(&pathtbl_resize_lock); if (grow) { set_bit(MESH_WORK_GROW_MPP_TABLE, &ifmsh->wrkq_flags); - ieee80211_queue_work(&local->hw, &ifmsh->work); + ieee80211_queue_work(&local->hw, &sdata->work); } return 0; diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 854ef4e..75b896d 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1653,7 +1653,7 @@ ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, case IEEE80211_STYPE_DISASSOC: case IEEE80211_STYPE_ACTION: skb_queue_tail(&sdata->skb_queue, skb); - ieee80211_queue_work(&local->hw, &sdata->u.mgd.work); + ieee80211_queue_work(&local->hw, &sdata->work); return RX_QUEUED; } @@ -1779,13 +1779,13 @@ static void ieee80211_sta_timer(unsigned long data) return; } - ieee80211_queue_work(&local->hw, &ifmgd->work); + ieee80211_queue_work(&local->hw, &sdata->work); } static void ieee80211_sta_work(struct work_struct *work) { struct ieee80211_sub_if_data *sdata = - container_of(work, struct ieee80211_sub_if_data, u.mgd.work); + container_of(work, struct ieee80211_sub_if_data, work); struct ieee80211_local *local = sdata->local; struct ieee80211_if_managed *ifmgd; struct sk_buff *skb; @@ -1906,8 +1906,7 @@ static void ieee80211_restart_sta_timer(struct ieee80211_sub_if_data *sdata) ieee80211_queue_work(&sdata->local->hw, &sdata->u.mgd.monitor_work); /* and do all the other regular work too */ - ieee80211_queue_work(&sdata->local->hw, - &sdata->u.mgd.work); + ieee80211_queue_work(&sdata->local->hw, &sdata->work); } } @@ -1922,7 +1921,6 @@ void ieee80211_sta_quiesce(struct ieee80211_sub_if_data *sdata) * time -- the code here is properly synchronised. */ - cancel_work_sync(&ifmgd->work); cancel_work_sync(&ifmgd->beacon_connection_loss_work); if (del_timer_sync(&ifmgd->timer)) set_bit(TMR_RUNNING_TIMER, &ifmgd->timers_running); @@ -1954,7 +1952,7 @@ void ieee80211_sta_setup_sdata(struct ieee80211_sub_if_data *sdata) struct ieee80211_if_managed *ifmgd; ifmgd = &sdata->u.mgd; - INIT_WORK(&ifmgd->work, ieee80211_sta_work); + INIT_WORK(&sdata->work, ieee80211_sta_work); INIT_WORK(&ifmgd->monitor_work, ieee80211_sta_monitor_work); INIT_WORK(&ifmgd->chswitch_work, ieee80211_chswitch_work); INIT_WORK(&ifmgd->beacon_connection_loss_work, diff --git a/net/mac80211/pm.c b/net/mac80211/pm.c index e145a94..d287fde 100644 --- a/net/mac80211/pm.c +++ b/net/mac80211/pm.c @@ -64,6 +64,8 @@ int __ieee80211_suspend(struct ieee80211_hw *hw) /* remove all interfaces */ list_for_each_entry(sdata, &local->interfaces, list) { + cancel_work_sync(&sdata->work); + switch(sdata->vif.type) { case NL80211_IFTYPE_STATION: ieee80211_sta_quiesce(sdata); -- cgit v1.1 From 1fa57d017366fb26b58af110a38b36a4f0214a62 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:32 +0200 Subject: mac80211: use common work function Even with the previous patch, IBSS, managed and mesh modes all attach their own work function to the shared work struct, which means some duplicated code. Change that to only have a frame processing function and a further work function for each of them and share some common code. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/ibss.c | 29 +++------------------- net/mac80211/ieee80211_i.h | 11 ++++++++ net/mac80211/iface.c | 62 ++++++++++++++++++++++++++++++++++++++++++++++ net/mac80211/mesh.c | 20 +++------------ net/mac80211/mlme.c | 35 +++----------------------- 5 files changed, 84 insertions(+), 73 deletions(-) (limited to 'net') diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index 18c4266..db5a479 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -727,8 +727,8 @@ static void ieee80211_rx_mgmt_beacon(struct ieee80211_sub_if_data *sdata, ieee80211_rx_bss_info(sdata, mgmt, len, rx_status, &elems, true); } -static void ieee80211_ibss_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, - struct sk_buff *skb) +void ieee80211_ibss_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, + struct sk_buff *skb) { struct ieee80211_rx_status *rx_status; struct ieee80211_mgmt *mgmt; @@ -758,29 +758,9 @@ static void ieee80211_ibss_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, kfree_skb(skb); } -static void ieee80211_ibss_work(struct work_struct *work) +void ieee80211_ibss_work(struct ieee80211_sub_if_data *sdata) { - struct ieee80211_sub_if_data *sdata = - container_of(work, struct ieee80211_sub_if_data, work); - struct ieee80211_local *local = sdata->local; - struct ieee80211_if_ibss *ifibss; - struct sk_buff *skb; - - if (WARN_ON(local->suspended)) - return; - - if (!ieee80211_sdata_running(sdata)) - return; - - if (local->scanning) - return; - - if (WARN_ON(sdata->vif.type != NL80211_IFTYPE_ADHOC)) - return; - ifibss = &sdata->u.ibss; - - while ((skb = skb_dequeue(&sdata->skb_queue))) - ieee80211_ibss_rx_queued_mgmt(sdata, skb); + struct ieee80211_if_ibss *ifibss = &sdata->u.ibss; if (!test_and_clear_bit(IEEE80211_IBSS_REQ_RUN, &ifibss->request)) return; @@ -846,7 +826,6 @@ void ieee80211_ibss_setup_sdata(struct ieee80211_sub_if_data *sdata) { struct ieee80211_if_ibss *ifibss = &sdata->u.ibss; - INIT_WORK(&sdata->work, ieee80211_ibss_work); setup_timer(&ifibss->timer, ieee80211_ibss_timer, (unsigned long) sdata); } diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 2873f63..f8c8f10 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -995,6 +995,9 @@ void ieee80211_sta_process_chanswitch(struct ieee80211_sub_if_data *sdata, u64 timestamp); void ieee80211_sta_quiesce(struct ieee80211_sub_if_data *sdata); void ieee80211_sta_restart(struct ieee80211_sub_if_data *sdata); +void ieee80211_sta_work(struct ieee80211_sub_if_data *sdata); +void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, + struct sk_buff *skb); /* IBSS code */ void ieee80211_ibss_notify_scan_completed(struct ieee80211_local *local); @@ -1009,6 +1012,14 @@ int ieee80211_ibss_join(struct ieee80211_sub_if_data *sdata, int ieee80211_ibss_leave(struct ieee80211_sub_if_data *sdata); void ieee80211_ibss_quiesce(struct ieee80211_sub_if_data *sdata); void ieee80211_ibss_restart(struct ieee80211_sub_if_data *sdata); +void ieee80211_ibss_work(struct ieee80211_sub_if_data *sdata); +void ieee80211_ibss_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, + struct sk_buff *skb); + +/* mesh code */ +void ieee80211_mesh_work(struct ieee80211_sub_if_data *sdata); +void ieee80211_mesh_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, + struct sk_buff *skb); /* scan/BSS handling */ void ieee80211_scan_work(struct work_struct *work); diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index de7ddc3..14212ad 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -701,6 +701,67 @@ static void ieee80211_if_setup(struct net_device *dev) dev->destructor = free_netdev; } +static void ieee80211_iface_work(struct work_struct *work) +{ + struct ieee80211_sub_if_data *sdata = + container_of(work, struct ieee80211_sub_if_data, work); + struct ieee80211_local *local = sdata->local; + struct sk_buff *skb; + + if (!ieee80211_sdata_running(sdata)) + return; + + if (local->scanning) + return; + + /* + * ieee80211_queue_work() should have picked up most cases, + * here we'll pick the rest. + */ + if (WARN(local->suspended, + "interface work scheduled while going to suspend\n")) + return; + + /* first process frames */ + while ((skb = skb_dequeue(&sdata->skb_queue))) { + switch (sdata->vif.type) { + case NL80211_IFTYPE_STATION: + ieee80211_sta_rx_queued_mgmt(sdata, skb); + break; + case NL80211_IFTYPE_ADHOC: + ieee80211_ibss_rx_queued_mgmt(sdata, skb); + break; + case NL80211_IFTYPE_MESH_POINT: + if (!ieee80211_vif_is_mesh(&sdata->vif)) + break; + ieee80211_mesh_rx_queued_mgmt(sdata, skb); + break; + default: + WARN(1, "frame for unexpected interface type"); + kfree_skb(skb); + break; + } + } + + /* then other type-dependent work */ + switch (sdata->vif.type) { + case NL80211_IFTYPE_STATION: + ieee80211_sta_work(sdata); + break; + case NL80211_IFTYPE_ADHOC: + ieee80211_ibss_work(sdata); + break; + case NL80211_IFTYPE_MESH_POINT: + if (!ieee80211_vif_is_mesh(&sdata->vif)) + break; + ieee80211_mesh_work(sdata); + break; + default: + break; + } +} + + /* * Helper function to initialise an interface to a specific type. */ @@ -719,6 +780,7 @@ static void ieee80211_setup_sdata(struct ieee80211_sub_if_data *sdata, sdata->dev->type = ARPHRD_ETHER; skb_queue_head_init(&sdata->skb_queue); + INIT_WORK(&sdata->work, ieee80211_iface_work); switch (type) { case NL80211_IFTYPE_AP: diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index be9aa98..0f1f593 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -596,8 +596,8 @@ static void ieee80211_mesh_rx_mgmt_action(struct ieee80211_sub_if_data *sdata, } } -static void ieee80211_mesh_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, - struct sk_buff *skb) +void ieee80211_mesh_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, + struct sk_buff *skb) { struct ieee80211_rx_status *rx_status; struct ieee80211_if_mesh *ifmsh; @@ -624,22 +624,9 @@ static void ieee80211_mesh_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, kfree_skb(skb); } -static void ieee80211_mesh_work(struct work_struct *work) +void ieee80211_mesh_work(struct ieee80211_sub_if_data *sdata) { - struct ieee80211_sub_if_data *sdata = - container_of(work, struct ieee80211_sub_if_data, work); - struct ieee80211_local *local = sdata->local; struct ieee80211_if_mesh *ifmsh = &sdata->u.mesh; - struct sk_buff *skb; - - if (!ieee80211_sdata_running(sdata)) - return; - - if (local->scanning) - return; - - while ((skb = skb_dequeue(&sdata->skb_queue))) - ieee80211_mesh_rx_queued_mgmt(sdata, skb); if (ifmsh->preq_queue_len && time_after(jiffies, @@ -674,7 +661,6 @@ void ieee80211_mesh_init_sdata(struct ieee80211_sub_if_data *sdata) { struct ieee80211_if_mesh *ifmsh = &sdata->u.mesh; - INIT_WORK(&sdata->work, ieee80211_mesh_work); setup_timer(&ifmsh->housekeeping_timer, ieee80211_mesh_housekeeping_timer, (unsigned long) sdata); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 75b896d..2f828ff 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1660,8 +1660,8 @@ ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, return RX_DROP_MONITOR; } -static void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, - struct sk_buff *skb) +void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, + struct sk_buff *skb) { struct ieee80211_if_managed *ifmgd = &sdata->u.mgd; struct ieee80211_rx_status *rx_status; @@ -1782,36 +1782,10 @@ static void ieee80211_sta_timer(unsigned long data) ieee80211_queue_work(&local->hw, &sdata->work); } -static void ieee80211_sta_work(struct work_struct *work) +void ieee80211_sta_work(struct ieee80211_sub_if_data *sdata) { - struct ieee80211_sub_if_data *sdata = - container_of(work, struct ieee80211_sub_if_data, work); struct ieee80211_local *local = sdata->local; - struct ieee80211_if_managed *ifmgd; - struct sk_buff *skb; - - if (!ieee80211_sdata_running(sdata)) - return; - - if (local->scanning) - return; - - if (WARN_ON(sdata->vif.type != NL80211_IFTYPE_STATION)) - return; - - /* - * ieee80211_queue_work() should have picked up most cases, - * here we'll pick the rest. - */ - if (WARN(local->suspended, "STA MLME work scheduled while " - "going to suspend\n")) - return; - - ifmgd = &sdata->u.mgd; - - /* first process frames to avoid timing out while a frame is pending */ - while ((skb = skb_dequeue(&sdata->skb_queue))) - ieee80211_sta_rx_queued_mgmt(sdata, skb); + struct ieee80211_if_managed *ifmgd = &sdata->u.mgd; /* then process the rest of the work */ mutex_lock(&ifmgd->mtx); @@ -1952,7 +1926,6 @@ void ieee80211_sta_setup_sdata(struct ieee80211_sub_if_data *sdata) struct ieee80211_if_managed *ifmgd; ifmgd = &sdata->u.mgd; - INIT_WORK(&sdata->work, ieee80211_sta_work); INIT_WORK(&ifmgd->monitor_work, ieee80211_sta_monitor_work); INIT_WORK(&ifmgd->chswitch_work, ieee80211_chswitch_work); INIT_WORK(&ifmgd->beacon_connection_loss_work, -- cgit v1.1 From 36b3a628a4e85d002ee8813ebd2a5caef6d3c1a7 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:33 +0200 Subject: mac80211: common work skb freeing All the management processing functions free the skb after they are done, so this can be done in the new common code instead. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/ibss.c | 2 -- net/mac80211/iface.c | 3 ++- net/mac80211/mesh.c | 2 -- net/mac80211/mlme.c | 5 +---- 4 files changed, 3 insertions(+), 9 deletions(-) (limited to 'net') diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index db5a479..982690a 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -754,8 +754,6 @@ void ieee80211_ibss_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, ieee80211_rx_mgmt_auth_ibss(sdata, mgmt, skb->len); break; } - - kfree_skb(skb); } void ieee80211_ibss_work(struct ieee80211_sub_if_data *sdata) diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 14212ad..1bf276d 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -738,9 +738,10 @@ static void ieee80211_iface_work(struct work_struct *work) break; default: WARN(1, "frame for unexpected interface type"); - kfree_skb(skb); break; } + + kfree_skb(skb); } /* then other type-dependent work */ diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index 0f1f593..83b346c 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -620,8 +620,6 @@ void ieee80211_mesh_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, ieee80211_mesh_rx_mgmt_action(sdata, mgmt, skb->len, rx_status); break; } - - kfree_skb(skb); } void ieee80211_mesh_work(struct ieee80211_sub_if_data *sdata) diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 2f828ff..4a5b29d 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1754,7 +1754,7 @@ void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, default: WARN(1, "unexpected: %d", rma); } - goto out; + return; } mutex_unlock(&ifmgd->mtx); @@ -1762,9 +1762,6 @@ void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, if (skb->len >= 24 + 2 /* mgmt + deauth reason */ && (fc & IEEE80211_FCTL_STYPE) == IEEE80211_STYPE_DEAUTH) cfg80211_send_deauth(sdata->dev, (u8 *)mgmt, skb->len); - - out: - kfree_skb(skb); } static void ieee80211_sta_timer(unsigned long data) -- cgit v1.1 From 77a121c3a88eb00a4b5e753d083dbb7d49fefb0a Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:34 +0200 Subject: mac80211: pull mgmt frame rx into rx handler Some code is duplicated between ibss, mesh and managed mode regarding the queueing of management frames. Since all modes now use a common skb queue and a common work function, we can pull the queueing code into the rx handler directly and remove the duplicated length checks etc. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/ibss.c | 26 -------------------- net/mac80211/ieee80211_i.h | 4 --- net/mac80211/mesh.c | 25 ------------------- net/mac80211/mesh.h | 2 -- net/mac80211/mlme.c | 27 -------------------- net/mac80211/rx.c | 61 ++++++++++++++++++++++++++++++++++++---------- 6 files changed, 48 insertions(+), 97 deletions(-) (limited to 'net') diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index 982690a..bfd72864 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -847,32 +847,6 @@ void ieee80211_ibss_notify_scan_completed(struct ieee80211_local *local) mutex_unlock(&local->iflist_mtx); } -ieee80211_rx_result -ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) -{ - struct ieee80211_local *local = sdata->local; - struct ieee80211_mgmt *mgmt; - u16 fc; - - if (skb->len < 24) - return RX_DROP_MONITOR; - - mgmt = (struct ieee80211_mgmt *) skb->data; - fc = le16_to_cpu(mgmt->frame_control); - - switch (fc & IEEE80211_FCTL_STYPE) { - case IEEE80211_STYPE_PROBE_RESP: - case IEEE80211_STYPE_BEACON: - case IEEE80211_STYPE_PROBE_REQ: - case IEEE80211_STYPE_AUTH: - skb_queue_tail(&sdata->skb_queue, skb); - ieee80211_queue_work(&local->hw, &sdata->work); - return RX_QUEUED; - } - - return RX_DROP_MONITOR; -} - int ieee80211_ibss_join(struct ieee80211_sub_if_data *sdata, struct cfg80211_ibss_params *params) { diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index f8c8f10..a3bcaa3 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -981,8 +981,6 @@ int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, int ieee80211_mgd_disassoc(struct ieee80211_sub_if_data *sdata, struct cfg80211_disassoc_request *req, void *cookie); -ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, - struct sk_buff *skb); void ieee80211_send_pspoll(struct ieee80211_local *local, struct ieee80211_sub_if_data *sdata); void ieee80211_recalc_ps(struct ieee80211_local *local, s32 latency); @@ -1002,8 +1000,6 @@ void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, /* IBSS code */ void ieee80211_ibss_notify_scan_completed(struct ieee80211_local *local); void ieee80211_ibss_setup_sdata(struct ieee80211_sub_if_data *sdata); -ieee80211_rx_result -ieee80211_ibss_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb); struct sta_info *ieee80211_ibss_add_sta(struct ieee80211_sub_if_data *sdata, u8 *bssid, u8 *addr, u32 supp_rates, gfp_t gfp); diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index 83b346c..c8a4f19 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -702,28 +702,3 @@ void ieee80211_mesh_init_sdata(struct ieee80211_sub_if_data *sdata) INIT_LIST_HEAD(&ifmsh->preq_queue.list); spin_lock_init(&ifmsh->mesh_preq_queue_lock); } - -ieee80211_rx_result -ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) -{ - struct ieee80211_local *local = sdata->local; - struct ieee80211_mgmt *mgmt; - u16 fc; - - if (skb->len < 24) - return RX_DROP_MONITOR; - - mgmt = (struct ieee80211_mgmt *) skb->data; - fc = le16_to_cpu(mgmt->frame_control); - - switch (fc & IEEE80211_FCTL_STYPE) { - case IEEE80211_STYPE_ACTION: - case IEEE80211_STYPE_PROBE_RESP: - case IEEE80211_STYPE_BEACON: - skb_queue_tail(&sdata->skb_queue, skb); - ieee80211_queue_work(&local->hw, &sdata->work); - return RX_QUEUED; - } - - return RX_CONTINUE; -} diff --git a/net/mac80211/mesh.h b/net/mac80211/mesh.h index c88087f..ebd3f1d 100644 --- a/net/mac80211/mesh.h +++ b/net/mac80211/mesh.h @@ -237,8 +237,6 @@ void ieee80211s_update_metric(struct ieee80211_local *local, struct sta_info *stainfo, struct sk_buff *skb); void ieee80211s_stop(void); void ieee80211_mesh_init_sdata(struct ieee80211_sub_if_data *sdata); -ieee80211_rx_result -ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb); void ieee80211_start_mesh(struct ieee80211_sub_if_data *sdata); void ieee80211_stop_mesh(struct ieee80211_sub_if_data *sdata); void ieee80211_mesh_root_setup(struct ieee80211_if_mesh *ifmsh); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 4a5b29d..036f1bf 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1633,33 +1633,6 @@ static void ieee80211_rx_mgmt_beacon(struct ieee80211_sub_if_data *sdata, ieee80211_bss_info_change_notify(sdata, changed); } -ieee80211_rx_result ieee80211_sta_rx_mgmt(struct ieee80211_sub_if_data *sdata, - struct sk_buff *skb) -{ - struct ieee80211_local *local = sdata->local; - struct ieee80211_mgmt *mgmt; - u16 fc; - - if (skb->len < 24) - return RX_DROP_MONITOR; - - mgmt = (struct ieee80211_mgmt *) skb->data; - fc = le16_to_cpu(mgmt->frame_control); - - switch (fc & IEEE80211_FCTL_STYPE) { - case IEEE80211_STYPE_PROBE_RESP: - case IEEE80211_STYPE_BEACON: - case IEEE80211_STYPE_DEAUTH: - case IEEE80211_STYPE_DISASSOC: - case IEEE80211_STYPE_ACTION: - skb_queue_tail(&sdata->skb_queue, skb); - ieee80211_queue_work(&local->hw, &sdata->work); - return RX_QUEUED; - } - - return RX_DROP_MONITOR; -} - void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) { diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 3adcda5..40fe279 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -1950,8 +1950,11 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) if (len < IEEE80211_MIN_ACTION_SIZE + 1) break; - if (sdata->vif.type == NL80211_IFTYPE_STATION) - return ieee80211_sta_rx_mgmt(sdata, rx->skb); + if (sdata->vif.type == NL80211_IFTYPE_STATION) { + skb_queue_tail(&sdata->skb_queue, rx->skb); + ieee80211_queue_work(&local->hw, &sdata->work); + return RX_QUEUED; + } switch (mgmt->u.action.u.addba_req.action_code) { case WLAN_ACTION_ADDBA_REQ: @@ -2003,7 +2006,9 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) if (memcmp(mgmt->bssid, sdata->u.mgd.bssid, ETH_ALEN)) break; - return ieee80211_sta_rx_mgmt(sdata, rx->skb); + skb_queue_tail(&sdata->skb_queue, rx->skb); + ieee80211_queue_work(&local->hw, &sdata->work); + return RX_QUEUED; } break; case WLAN_CATEGORY_SA_QUERY: @@ -2021,9 +2026,11 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) break; case WLAN_CATEGORY_MESH_PLINK: case WLAN_CATEGORY_MESH_PATH_SEL: - if (ieee80211_vif_is_mesh(&sdata->vif)) - return ieee80211_mesh_rx_mgmt(sdata, rx->skb); - break; + if (!ieee80211_vif_is_mesh(&sdata->vif)) + break; + skb_queue_tail(&sdata->skb_queue, rx->skb); + ieee80211_queue_work(&local->hw, &sdata->work); + return RX_QUEUED; } /* @@ -2080,10 +2087,15 @@ ieee80211_rx_h_mgmt(struct ieee80211_rx_data *rx) { struct ieee80211_sub_if_data *sdata = rx->sdata; ieee80211_rx_result rxs; + struct ieee80211_mgmt *mgmt = (void *)rx->skb->data; + __le16 stype; if (!(rx->flags & IEEE80211_RX_RA_MATCH)) return RX_DROP_MONITOR; + if (rx->skb->len < 24) + return RX_DROP_MONITOR; + if (ieee80211_drop_unencrypted_mgmt(rx)) return RX_DROP_UNUSABLE; @@ -2091,16 +2103,39 @@ ieee80211_rx_h_mgmt(struct ieee80211_rx_data *rx) if (rxs != RX_CONTINUE) return rxs; - if (ieee80211_vif_is_mesh(&sdata->vif)) - return ieee80211_mesh_rx_mgmt(sdata, rx->skb); + stype = mgmt->frame_control & cpu_to_le16(IEEE80211_FCTL_STYPE); - if (sdata->vif.type == NL80211_IFTYPE_ADHOC) - return ieee80211_ibss_rx_mgmt(sdata, rx->skb); + if (!ieee80211_vif_is_mesh(&sdata->vif) && + sdata->vif.type != NL80211_IFTYPE_ADHOC && + sdata->vif.type != NL80211_IFTYPE_STATION) + return RX_DROP_MONITOR; - if (sdata->vif.type == NL80211_IFTYPE_STATION) - return ieee80211_sta_rx_mgmt(sdata, rx->skb); + switch (stype) { + case cpu_to_le16(IEEE80211_STYPE_BEACON): + case cpu_to_le16(IEEE80211_STYPE_PROBE_RESP): + /* process for all: mesh, mlme, ibss */ + break; + case cpu_to_le16(IEEE80211_STYPE_DEAUTH): + case cpu_to_le16(IEEE80211_STYPE_DISASSOC): + /* process only for station */ + if (sdata->vif.type != NL80211_IFTYPE_STATION) + return RX_DROP_MONITOR; + break; + case cpu_to_le16(IEEE80211_STYPE_PROBE_REQ): + case cpu_to_le16(IEEE80211_STYPE_AUTH): + /* process only for ibss */ + if (sdata->vif.type != NL80211_IFTYPE_ADHOC) + return RX_DROP_MONITOR; + break; + default: + return RX_DROP_MONITOR; + } - return RX_DROP_MONITOR; + /* queue up frame and kick off work to process it */ + skb_queue_tail(&sdata->skb_queue, rx->skb); + ieee80211_queue_work(&rx->local->hw, &sdata->work); + + return RX_QUEUED; } static void ieee80211_rx_michael_mic_report(struct ieee80211_hdr *hdr, -- cgit v1.1 From bed7ee6e44cb7633a4f9821688a6c7ae977615ed Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:35 +0200 Subject: mac80211: always process blockack action from workqueue To prepare for making the ampdu_action callback sleep, make mac80211 always process blockack action frames from the skb queue. This gets rid of the current special case for managed mode interfaces as well. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/iface.c | 31 ++++++++++++++++++++++++++++++- net/mac80211/mlme.c | 38 -------------------------------------- net/mac80211/rx.c | 29 +++++++++++++---------------- 3 files changed, 43 insertions(+), 55 deletions(-) (limited to 'net') diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 1bf276d..7a3dbde 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -724,7 +724,36 @@ static void ieee80211_iface_work(struct work_struct *work) /* first process frames */ while ((skb = skb_dequeue(&sdata->skb_queue))) { - switch (sdata->vif.type) { + struct ieee80211_mgmt *mgmt = (void *)skb->data; + + if (ieee80211_is_action(mgmt->frame_control) && + mgmt->u.action.category == WLAN_CATEGORY_BACK) { + int len = skb->len; + struct sta_info *sta; + + rcu_read_lock(); + sta = sta_info_get(sdata, mgmt->sa); + if (sta) { + switch (mgmt->u.action.u.addba_req.action_code) { + case WLAN_ACTION_ADDBA_REQ: + ieee80211_process_addba_request( + local, sta, mgmt, len); + break; + case WLAN_ACTION_ADDBA_RESP: + ieee80211_process_addba_resp(local, sta, + mgmt, len); + break; + case WLAN_ACTION_DELBA: + ieee80211_process_delba(sdata, sta, + mgmt, len); + break; + default: + WARN_ON(1); + break; + } + } + rcu_read_unlock(); + } else switch (sdata->vif.type) { case NL80211_IFTYPE_STATION: ieee80211_sta_rx_queued_mgmt(sdata, skb); break; diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 036f1bf..583b346 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1666,44 +1666,6 @@ void ieee80211_sta_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata, break; case IEEE80211_STYPE_ACTION: switch (mgmt->u.action.category) { - case WLAN_CATEGORY_BACK: { - struct ieee80211_local *local = sdata->local; - int len = skb->len; - struct sta_info *sta; - - rcu_read_lock(); - sta = sta_info_get(sdata, mgmt->sa); - if (!sta) { - rcu_read_unlock(); - break; - } - - local_bh_disable(); - - switch (mgmt->u.action.u.addba_req.action_code) { - case WLAN_ACTION_ADDBA_REQ: - if (len < (IEEE80211_MIN_ACTION_SIZE + - sizeof(mgmt->u.action.u.addba_req))) - break; - ieee80211_process_addba_request(local, sta, mgmt, len); - break; - case WLAN_ACTION_ADDBA_RESP: - if (len < (IEEE80211_MIN_ACTION_SIZE + - sizeof(mgmt->u.action.u.addba_resp))) - break; - ieee80211_process_addba_resp(local, sta, mgmt, len); - break; - case WLAN_ACTION_DELBA: - if (len < (IEEE80211_MIN_ACTION_SIZE + - sizeof(mgmt->u.action.u.delba))) - break; - ieee80211_process_delba(sdata, sta, mgmt, len); - break; - } - local_bh_enable(); - rcu_read_unlock(); - break; - } case WLAN_CATEGORY_SPECTRUM_MGMT: ieee80211_sta_process_chanswitch(sdata, &mgmt->u.action.u.chan_switch.sw_elem, diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 40fe279..435c216 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -1950,33 +1950,29 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) if (len < IEEE80211_MIN_ACTION_SIZE + 1) break; - if (sdata->vif.type == NL80211_IFTYPE_STATION) { - skb_queue_tail(&sdata->skb_queue, rx->skb); - ieee80211_queue_work(&local->hw, &sdata->work); - return RX_QUEUED; - } - switch (mgmt->u.action.u.addba_req.action_code) { case WLAN_ACTION_ADDBA_REQ: if (len < (IEEE80211_MIN_ACTION_SIZE + sizeof(mgmt->u.action.u.addba_req))) - return RX_DROP_MONITOR; - ieee80211_process_addba_request(local, rx->sta, mgmt, len); - goto handled; + goto invalid; + break; case WLAN_ACTION_ADDBA_RESP: if (len < (IEEE80211_MIN_ACTION_SIZE + sizeof(mgmt->u.action.u.addba_resp))) - break; - ieee80211_process_addba_resp(local, rx->sta, mgmt, len); - goto handled; + goto invalid; + break; case WLAN_ACTION_DELBA: if (len < (IEEE80211_MIN_ACTION_SIZE + sizeof(mgmt->u.action.u.delba))) - break; - ieee80211_process_delba(sdata, rx->sta, mgmt, len); - goto handled; + goto invalid; + break; + default: + goto invalid; } - break; + + skb_queue_tail(&sdata->skb_queue, rx->skb); + ieee80211_queue_work(&local->hw, &sdata->work); + return RX_QUEUED; case WLAN_CATEGORY_SPECTRUM_MGMT: if (local->hw.conf.channel->band != IEEE80211_BAND_5GHZ) break; @@ -2033,6 +2029,7 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) return RX_QUEUED; } + invalid: /* * For AP mode, hostapd is responsible for handling any action * frames that we didn't handle, including returning unknown -- cgit v1.1 From 344eec67c7b8557234e149d254bca2ae9614d61e Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:36 +0200 Subject: mac80211: move blockack stop due to fragmentation There's a corner case where we receive a fragmented frame during a blockack session, in which case we will terminate that session. To simplify future work in this area that will culminate in allowing the driver callbacks for aggregation to sleep, move the processing of this case out of the RX path into the interface work. This will simplify future work because the new place for this code doesn't require that the function will always be atomic, which the RX path needs. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/iface.c | 32 +++++++++++++++++++++++++++++++- net/mac80211/rx.c | 5 ++--- 2 files changed, 33 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 7a3dbde..87fc012 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -707,6 +707,7 @@ static void ieee80211_iface_work(struct work_struct *work) container_of(work, struct ieee80211_sub_if_data, work); struct ieee80211_local *local = sdata->local; struct sk_buff *skb; + struct sta_info *sta; if (!ieee80211_sdata_running(sdata)) return; @@ -729,7 +730,6 @@ static void ieee80211_iface_work(struct work_struct *work) if (ieee80211_is_action(mgmt->frame_control) && mgmt->u.action.category == WLAN_CATEGORY_BACK) { int len = skb->len; - struct sta_info *sta; rcu_read_lock(); sta = sta_info_get(sdata, mgmt->sa); @@ -753,6 +753,36 @@ static void ieee80211_iface_work(struct work_struct *work) } } rcu_read_unlock(); + } else if (ieee80211_is_data_qos(mgmt->frame_control)) { + struct ieee80211_hdr *hdr = (void *)mgmt; + /* + * So the frame isn't mgmt, but frame_control + * is at the right place anyway, of course, so + * the if statement is correct. + * + * Warn if we have other data frame types here, + * they must not get here. + */ + WARN_ON(hdr->frame_control & + cpu_to_le16(IEEE80211_STYPE_NULLFUNC)); + WARN_ON(!(hdr->seq_ctrl & + cpu_to_le16(IEEE80211_SCTL_FRAG))); + /* + * This was a fragment of a frame, received while + * a block-ack session was active. That cannot be + * right, so terminate the session. + */ + rcu_read_lock(); + sta = sta_info_get(sdata, mgmt->sa); + if (sta) { + u16 tid = *ieee80211_get_qos_ctl(hdr) & + IEEE80211_QOS_CTL_TID_MASK; + + __ieee80211_stop_rx_ba_session( + sta, tid, WLAN_BACK_RECIPIENT, + WLAN_REASON_QSTA_REQUIRE_SETUP); + } + rcu_read_unlock(); } else switch (sdata->vif.type) { case NL80211_IFTYPE_STATION: ieee80211_sta_rx_queued_mgmt(sdata, skb); diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 435c216..b716fa2 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -741,9 +741,8 @@ static void ieee80211_rx_reorder_ampdu(struct ieee80211_rx_data *rx, sc = le16_to_cpu(hdr->seq_ctrl); if (sc & IEEE80211_SCTL_FRAG) { spin_unlock(&sta->lock); - __ieee80211_stop_rx_ba_session(sta, tid, WLAN_BACK_RECIPIENT, - WLAN_REASON_QSTA_REQUIRE_SETUP); - dev_kfree_skb(skb); + skb_queue_tail(&rx->sdata->skb_queue, skb); + ieee80211_queue_work(&local->hw, &rx->sdata->work); return; } -- cgit v1.1 From c1475ca99edcc7216ddc45838ab2c3281c14ba22 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:37 +0200 Subject: mac80211: move aggregation callback processing This moves the aggregation callback processing to the per-sdata skb queue and a work function rather than the tasklet. Unfortunately, this means that it extends the pkt_type hack to that skb queue. However, it will enable making ampdu_action API changes gradually, my current plan is to get rid of this again by forcing drivers to only return from ampdu_action() when everything is done, thus removing the callbacks completely. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 14 ++++++-------- net/mac80211/ieee80211_i.h | 11 +++++++---- net/mac80211/iface.c | 15 ++++++++++++--- net/mac80211/main.c | 13 ------------- net/mac80211/rx.c | 5 +++++ 5 files changed, 30 insertions(+), 28 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index 9b9f21b..c7b7ac4 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -502,11 +502,10 @@ void ieee80211_start_tx_ba_cb_irqsafe(struct ieee80211_vif *vif, ra_tid = (struct ieee80211_ra_tid *) &skb->cb; memcpy(&ra_tid->ra, ra, ETH_ALEN); ra_tid->tid = tid; - ra_tid->vif = vif; - skb->pkt_type = IEEE80211_ADDBA_MSG; - skb_queue_tail(&local->skb_queue, skb); - tasklet_schedule(&local->tasklet); + skb->pkt_type = IEEE80211_SDATA_QUEUE_AGG_START; + skb_queue_tail(&sdata->skb_queue, skb); + ieee80211_queue_work(&local->hw, &sdata->work); } EXPORT_SYMBOL(ieee80211_start_tx_ba_cb_irqsafe); @@ -637,11 +636,10 @@ void ieee80211_stop_tx_ba_cb_irqsafe(struct ieee80211_vif *vif, ra_tid = (struct ieee80211_ra_tid *) &skb->cb; memcpy(&ra_tid->ra, ra, ETH_ALEN); ra_tid->tid = tid; - ra_tid->vif = vif; - skb->pkt_type = IEEE80211_DELBA_MSG; - skb_queue_tail(&local->skb_queue, skb); - tasklet_schedule(&local->tasklet); + skb->pkt_type = IEEE80211_SDATA_QUEUE_AGG_STOP; + skb_queue_tail(&sdata->skb_queue, skb); + ieee80211_queue_work(&local->hw, &sdata->work); } EXPORT_SYMBOL(ieee80211_stop_tx_ba_cb_irqsafe); diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index a3bcaa3..bafe610 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -564,11 +564,15 @@ ieee80211_sdata_set_mesh_id(struct ieee80211_sub_if_data *sdata, #endif } +enum sdata_queue_type { + IEEE80211_SDATA_QUEUE_TYPE_FRAME = 0, + IEEE80211_SDATA_QUEUE_AGG_START = 1, + IEEE80211_SDATA_QUEUE_AGG_STOP = 2, +}; + enum { IEEE80211_RX_MSG = 1, IEEE80211_TX_STATUS_MSG = 2, - IEEE80211_DELBA_MSG = 3, - IEEE80211_ADDBA_MSG = 4, }; enum queue_stop_reason { @@ -870,9 +874,8 @@ IEEE80211_DEV_TO_SUB_IF(struct net_device *dev) return netdev_priv(dev); } -/* this struct represents 802.11n's RA/TID combination along with our vif */ +/* this struct represents 802.11n's RA/TID combination */ struct ieee80211_ra_tid { - struct ieee80211_vif *vif; u8 ra[ETH_ALEN]; u16 tid; }; diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 87fc012..9f00d3f 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -482,7 +482,7 @@ static int ieee80211_stop(struct net_device *dev) } /* fall through */ default: - cancel_work_sync(&sdata->work); + flush_work(&sdata->work); /* * When we get here, the interface is marked down. * Call synchronize_rcu() to wait for the RX path @@ -708,6 +708,7 @@ static void ieee80211_iface_work(struct work_struct *work) struct ieee80211_local *local = sdata->local; struct sk_buff *skb; struct sta_info *sta; + struct ieee80211_ra_tid *ra_tid; if (!ieee80211_sdata_running(sdata)) return; @@ -727,8 +728,16 @@ static void ieee80211_iface_work(struct work_struct *work) while ((skb = skb_dequeue(&sdata->skb_queue))) { struct ieee80211_mgmt *mgmt = (void *)skb->data; - if (ieee80211_is_action(mgmt->frame_control) && - mgmt->u.action.category == WLAN_CATEGORY_BACK) { + if (skb->pkt_type == IEEE80211_SDATA_QUEUE_AGG_START) { + ra_tid = (void *)&skb->cb; + ieee80211_start_tx_ba_cb(&sdata->vif, ra_tid->ra, + ra_tid->tid); + } else if (skb->pkt_type == IEEE80211_SDATA_QUEUE_AGG_STOP) { + ra_tid = (void *)&skb->cb; + ieee80211_stop_tx_ba_cb(&sdata->vif, ra_tid->ra, + ra_tid->tid); + } else if (ieee80211_is_action(mgmt->frame_control) && + mgmt->u.action.category == WLAN_CATEGORY_BACK) { int len = skb->len; rcu_read_lock(); diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 88b671a..a2d10d4 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -259,7 +259,6 @@ static void ieee80211_tasklet_handler(unsigned long data) { struct ieee80211_local *local = (struct ieee80211_local *) data; struct sk_buff *skb; - struct ieee80211_ra_tid *ra_tid; while ((skb = skb_dequeue(&local->skb_queue)) || (skb = skb_dequeue(&local->skb_queue_unreliable))) { @@ -274,18 +273,6 @@ static void ieee80211_tasklet_handler(unsigned long data) skb->pkt_type = 0; ieee80211_tx_status(local_to_hw(local), skb); break; - case IEEE80211_DELBA_MSG: - ra_tid = (struct ieee80211_ra_tid *) &skb->cb; - ieee80211_stop_tx_ba_cb(ra_tid->vif, ra_tid->ra, - ra_tid->tid); - dev_kfree_skb(skb); - break; - case IEEE80211_ADDBA_MSG: - ra_tid = (struct ieee80211_ra_tid *) &skb->cb; - ieee80211_start_tx_ba_cb(ra_tid->vif, ra_tid->ra, - ra_tid->tid); - dev_kfree_skb(skb); - break ; default: WARN(1, "mac80211: Packet is of unknown type %d\n", skb->pkt_type); diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index b716fa2..84f1173 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -741,6 +741,7 @@ static void ieee80211_rx_reorder_ampdu(struct ieee80211_rx_data *rx, sc = le16_to_cpu(hdr->seq_ctrl); if (sc & IEEE80211_SCTL_FRAG) { spin_unlock(&sta->lock); + skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; skb_queue_tail(&rx->sdata->skb_queue, skb); ieee80211_queue_work(&local->hw, &rx->sdata->work); return; @@ -1969,6 +1970,7 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) goto invalid; } + rx->skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; skb_queue_tail(&sdata->skb_queue, rx->skb); ieee80211_queue_work(&local->hw, &sdata->work); return RX_QUEUED; @@ -2001,6 +2003,7 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) if (memcmp(mgmt->bssid, sdata->u.mgd.bssid, ETH_ALEN)) break; + rx->skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; skb_queue_tail(&sdata->skb_queue, rx->skb); ieee80211_queue_work(&local->hw, &sdata->work); return RX_QUEUED; @@ -2023,6 +2026,7 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) case WLAN_CATEGORY_MESH_PATH_SEL: if (!ieee80211_vif_is_mesh(&sdata->vif)) break; + rx->skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; skb_queue_tail(&sdata->skb_queue, rx->skb); ieee80211_queue_work(&local->hw, &sdata->work); return RX_QUEUED; @@ -2128,6 +2132,7 @@ ieee80211_rx_h_mgmt(struct ieee80211_rx_data *rx) } /* queue up frame and kick off work to process it */ + rx->skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; skb_queue_tail(&sdata->skb_queue, rx->skb); ieee80211_queue_work(&rx->local->hw, &sdata->work); -- cgit v1.1 From a87f736d942c86255e3088c606f0e3eab6bbf784 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:38 +0200 Subject: mac80211: use RCU for RX aggregation Currently we allocate some memory for each RX aggregation session and additionally keep a flag indicating whether or not it is valid. By using RCU to protect the pointer and making sure that the memory is fully set up before it becomes visible to the RX path, we can remove the need for the bool that indicates validity, as well as for locking on the RX path since it is always synchronised against itself, and we can guarantee that all other modifications are done when the structure is not visible to the RX path. The net result is that since we remove locking requirements from the RX path, we can in the future use any kind of lock for the setup and teardown code paths. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-rx.c | 70 +++++++++++++++++++++------------------------- net/mac80211/debugfs_sta.c | 6 ++-- net/mac80211/rx.c | 34 ++++++++++------------ net/mac80211/sta_info.h | 14 ++++++++-- 4 files changed, 61 insertions(+), 63 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-rx.c b/net/mac80211/agg-rx.c index 6bb9a9a..bbf36d9 100644 --- a/net/mac80211/agg-rx.c +++ b/net/mac80211/agg-rx.c @@ -19,25 +19,36 @@ #include "ieee80211_i.h" #include "driver-ops.h" +static void ieee80211_free_tid_rx(struct rcu_head *h) +{ + struct tid_ampdu_rx *tid_rx = + container_of(h, struct tid_ampdu_rx, rcu_head); + int i; + + for (i = 0; i < tid_rx->buf_size; i++) + dev_kfree_skb(tid_rx->reorder_buf[i]); + kfree(tid_rx->reorder_buf); + kfree(tid_rx->reorder_time); + kfree(tid_rx); +} + static void ___ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, u16 initiator, u16 reason, bool from_timer) { struct ieee80211_local *local = sta->local; struct tid_ampdu_rx *tid_rx; - int i; spin_lock_bh(&sta->lock); - /* check if TID is in operational state */ - if (!sta->ampdu_mlme.tid_active_rx[tid]) { + tid_rx = sta->ampdu_mlme.tid_rx[tid]; + + if (!tid_rx) { spin_unlock_bh(&sta->lock); return; } - sta->ampdu_mlme.tid_active_rx[tid] = false; - - tid_rx = sta->ampdu_mlme.tid_rx[tid]; + rcu_assign_pointer(sta->ampdu_mlme.tid_rx[tid], NULL); #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "Rx BA session stop requested for %pM tid %u\n", @@ -54,26 +65,12 @@ static void ___ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, ieee80211_send_delba(sta->sdata, sta->sta.addr, tid, 0, reason); - /* free the reordering buffer */ - for (i = 0; i < tid_rx->buf_size; i++) { - if (tid_rx->reorder_buf[i]) { - /* release the reordered frames */ - dev_kfree_skb(tid_rx->reorder_buf[i]); - tid_rx->stored_mpdu_num--; - tid_rx->reorder_buf[i] = NULL; - } - } - - /* free resources */ - kfree(tid_rx->reorder_buf); - kfree(tid_rx->reorder_time); - sta->ampdu_mlme.tid_rx[tid] = NULL; - spin_unlock_bh(&sta->lock); if (!from_timer) del_timer_sync(&tid_rx->session_timer); - kfree(tid_rx); + + call_rcu(&tid_rx->rcu_head, ieee80211_free_tid_rx); } void __ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, @@ -214,7 +211,7 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, /* examine state machine */ spin_lock_bh(&sta->lock); - if (sta->ampdu_mlme.tid_active_rx[tid]) { + if (sta->ampdu_mlme.tid_rx[tid]) { #ifdef CONFIG_MAC80211_HT_DEBUG if (net_ratelimit()) printk(KERN_DEBUG "unexpected AddBA Req from " @@ -225,9 +222,8 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, } /* prepare A-MPDU MLME for Rx aggregation */ - sta->ampdu_mlme.tid_rx[tid] = - kmalloc(sizeof(struct tid_ampdu_rx), GFP_ATOMIC); - if (!sta->ampdu_mlme.tid_rx[tid]) { + tid_agg_rx = kmalloc(sizeof(struct tid_ampdu_rx), GFP_ATOMIC); + if (!tid_agg_rx) { #ifdef CONFIG_MAC80211_HT_DEBUG if (net_ratelimit()) printk(KERN_ERR "allocate rx mlme to tid %d failed\n", @@ -235,14 +231,11 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, #endif goto end; } - /* rx timer */ - sta->ampdu_mlme.tid_rx[tid]->session_timer.function = - sta_rx_agg_session_timer_expired; - sta->ampdu_mlme.tid_rx[tid]->session_timer.data = - (unsigned long)&sta->timer_to_tid[tid]; - init_timer(&sta->ampdu_mlme.tid_rx[tid]->session_timer); - tid_agg_rx = sta->ampdu_mlme.tid_rx[tid]; + /* rx timer */ + tid_agg_rx->session_timer.function = sta_rx_agg_session_timer_expired; + tid_agg_rx->session_timer.data = (unsigned long)&sta->timer_to_tid[tid]; + init_timer(&tid_agg_rx->session_timer); /* prepare reordering buffer */ tid_agg_rx->reorder_buf = @@ -257,8 +250,7 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, #endif kfree(tid_agg_rx->reorder_buf); kfree(tid_agg_rx->reorder_time); - kfree(sta->ampdu_mlme.tid_rx[tid]); - sta->ampdu_mlme.tid_rx[tid] = NULL; + kfree(tid_agg_rx); goto end; } @@ -270,13 +262,12 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, if (ret) { kfree(tid_agg_rx->reorder_buf); + kfree(tid_agg_rx->reorder_time); kfree(tid_agg_rx); - sta->ampdu_mlme.tid_rx[tid] = NULL; goto end; } - /* change state and send addba resp */ - sta->ampdu_mlme.tid_active_rx[tid] = true; + /* update data */ tid_agg_rx->dialog_token = dialog_token; tid_agg_rx->ssn = start_seq_num; tid_agg_rx->head_seq_num = start_seq_num; @@ -284,6 +275,9 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, tid_agg_rx->timeout = timeout; tid_agg_rx->stored_mpdu_num = 0; status = WLAN_STATUS_SUCCESS; + + /* activate it for RX */ + rcu_assign_pointer(sta->ampdu_mlme.tid_rx[tid], tid_agg_rx); end: spin_unlock_bh(&sta->lock); diff --git a/net/mac80211/debugfs_sta.c b/net/mac80211/debugfs_sta.c index 576e024..95c1ea4 100644 --- a/net/mac80211/debugfs_sta.c +++ b/net/mac80211/debugfs_sta.c @@ -125,12 +125,12 @@ static ssize_t sta_agg_status_read(struct file *file, char __user *userbuf, for (i = 0; i < STA_TID_NUM; i++) { p += scnprintf(p, sizeof(buf) + buf - p, "%02d", i); p += scnprintf(p, sizeof(buf) + buf - p, "\t\t%x", - sta->ampdu_mlme.tid_active_rx[i]); + !!sta->ampdu_mlme.tid_rx[i]); p += scnprintf(p, sizeof(buf) + buf - p, "\t%#.2x", - sta->ampdu_mlme.tid_active_rx[i] ? + sta->ampdu_mlme.tid_rx[i] ? sta->ampdu_mlme.tid_rx[i]->dialog_token : 0); p += scnprintf(p, sizeof(buf) + buf - p, "\t%#.3x", - sta->ampdu_mlme.tid_active_rx[i] ? + sta->ampdu_mlme.tid_rx[i] ? sta->ampdu_mlme.tid_rx[i]->ssn : 0); p += scnprintf(p, sizeof(buf) + buf - p, "\t\t%x", diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 84f1173..ee01dac 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -719,16 +719,13 @@ static void ieee80211_rx_reorder_ampdu(struct ieee80211_rx_data *rx, tid = *ieee80211_get_qos_ctl(hdr) & IEEE80211_QOS_CTL_TID_MASK; - spin_lock(&sta->lock); - - if (!sta->ampdu_mlme.tid_active_rx[tid]) - goto dont_reorder_unlock; - - tid_agg_rx = sta->ampdu_mlme.tid_rx[tid]; + tid_agg_rx = rcu_dereference(sta->ampdu_mlme.tid_rx[tid]); + if (!tid_agg_rx) + goto dont_reorder; /* qos null data frames are excluded */ if (unlikely(hdr->frame_control & cpu_to_le16(IEEE80211_STYPE_NULLFUNC))) - goto dont_reorder_unlock; + goto dont_reorder; /* new, potentially un-ordered, ampdu frame - process it */ @@ -740,20 +737,22 @@ static void ieee80211_rx_reorder_ampdu(struct ieee80211_rx_data *rx, /* if this mpdu is fragmented - terminate rx aggregation session */ sc = le16_to_cpu(hdr->seq_ctrl); if (sc & IEEE80211_SCTL_FRAG) { - spin_unlock(&sta->lock); skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; skb_queue_tail(&rx->sdata->skb_queue, skb); ieee80211_queue_work(&local->hw, &rx->sdata->work); return; } - if (ieee80211_sta_manage_reorder_buf(hw, tid_agg_rx, skb, frames)) { - spin_unlock(&sta->lock); + /* + * No locking needed -- we will only ever process one + * RX packet at a time, and thus own tid_agg_rx. All + * other code manipulating it needs to (and does) make + * sure that we cannot get to it any more before doing + * anything with it. + */ + if (ieee80211_sta_manage_reorder_buf(hw, tid_agg_rx, skb, frames)) return; - } - dont_reorder_unlock: - spin_unlock(&sta->lock); dont_reorder: __skb_queue_tail(frames, skb); } @@ -1830,13 +1829,11 @@ ieee80211_rx_h_ctrl(struct ieee80211_rx_data *rx, struct sk_buff_head *frames) &bar_data, sizeof(bar_data))) return RX_DROP_MONITOR; - spin_lock(&rx->sta->lock); tid = le16_to_cpu(bar_data.control) >> 12; - if (!rx->sta->ampdu_mlme.tid_active_rx[tid]) { - spin_unlock(&rx->sta->lock); + + tid_agg_rx = rcu_dereference(rx->sta->ampdu_mlme.tid_rx[tid]); + if (!tid_agg_rx) return RX_DROP_MONITOR; - } - tid_agg_rx = rx->sta->ampdu_mlme.tid_rx[tid]; start_seq_num = le16_to_cpu(bar_data.start_seq_num) >> 4; @@ -1849,7 +1846,6 @@ ieee80211_rx_h_ctrl(struct ieee80211_rx_data *rx, struct sk_buff_head *frames) ieee80211_release_reorder_frames(hw, tid_agg_rx, start_seq_num, frames); kfree_skb(skb); - spin_unlock(&rx->sta->lock); return RX_QUEUED; } diff --git a/net/mac80211/sta_info.h b/net/mac80211/sta_info.h index 786bbd3..16864a6 100644 --- a/net/mac80211/sta_info.h +++ b/net/mac80211/sta_info.h @@ -102,8 +102,18 @@ struct tid_ampdu_tx { * @buf_size: buffer size for incoming A-MPDUs * @timeout: reset timer value (in TUs). * @dialog_token: dialog token for aggregation session + * @rcu_head: RCU head used for freeing this struct + * + * This structure is protected by RCU and the per-station + * spinlock. Assignments to the array holding it must hold + * the spinlock, only the RX path can access it under RCU + * lock-free. The RX path, since it is single-threaded, + * can even modify the structure without locking since the + * only other modifications to it are done when the struct + * can not yet or no longer be found by the RX path. */ struct tid_ampdu_rx { + struct rcu_head rcu_head; struct sk_buff **reorder_buf; unsigned long *reorder_time; struct timer_list session_timer; @@ -118,8 +128,7 @@ struct tid_ampdu_rx { /** * struct sta_ampdu_mlme - STA aggregation information. * - * @tid_active_rx: TID's state in Rx session state machine. - * @tid_rx: aggregation info for Rx per TID + * @tid_rx: aggregation info for Rx per TID -- RCU protected * @tid_state_tx: TID's state in Tx session state machine. * @tid_tx: aggregation info for Tx per TID * @addba_req_num: number of times addBA request has been sent. @@ -127,7 +136,6 @@ struct tid_ampdu_rx { */ struct sta_ampdu_mlme { /* rx */ - bool tid_active_rx[STA_TID_NUM]; struct tid_ampdu_rx *tid_rx[STA_TID_NUM]; /* tx */ u8 tid_state_tx[STA_TID_NUM]; -- cgit v1.1 From a622ab72b4dcfdf53e24b16e9530cb876979a00c Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:39 +0200 Subject: mac80211: use RCU for TX aggregation Currently we allocate some memory for each TX aggregation session and additionally keep a state bitmap indicating the state it is in. By using RCU to protect the pointer, moving the state into the structure and some locking trickery we can avoid locking when the TX agg session is fully operational. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 229 +++++++++++++++++++------------------ net/mac80211/debugfs_sta.c | 8 +- net/mac80211/ht.c | 9 +- net/mac80211/ieee80211_i.h | 2 - net/mac80211/rc80211_minstrel_ht.c | 2 +- net/mac80211/sta_info.c | 12 +- net/mac80211/sta_info.h | 33 +++--- net/mac80211/tx.c | 88 +++++++++----- 8 files changed, 206 insertions(+), 177 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index c7b7ac4..7d8656d 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -125,25 +125,42 @@ void ieee80211_send_bar(struct ieee80211_sub_if_data *sdata, u8 *ra, u16 tid, u1 ieee80211_tx_skb(sdata, skb); } -int ___ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, - enum ieee80211_back_parties initiator) +static void kfree_tid_tx(struct rcu_head *rcu_head) +{ + struct tid_ampdu_tx *tid_tx = + container_of(rcu_head, struct tid_ampdu_tx, rcu_head); + + kfree(tid_tx); +} + +static int ___ieee80211_stop_tx_ba_session( + struct sta_info *sta, u16 tid, + enum ieee80211_back_parties initiator) { struct ieee80211_local *local = sta->local; + struct tid_ampdu_tx *tid_tx = sta->ampdu_mlme.tid_tx[tid]; int ret; - u8 *state; + + lockdep_assert_held(&sta->lock); + + if (WARN_ON(!tid_tx)) + return -ENOENT; #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "Tx BA session stop requested for %pM tid %u\n", sta->sta.addr, tid); #endif /* CONFIG_MAC80211_HT_DEBUG */ - state = &sta->ampdu_mlme.tid_state_tx[tid]; + set_bit(HT_AGG_STATE_STOPPING, &tid_tx->state); - if (*state == HT_AGG_STATE_OPERATIONAL) - sta->ampdu_mlme.addba_req_num[tid] = 0; + /* + * After this packets are no longer handed right through + * to the driver but are put onto tid_tx->pending instead, + * with locking to ensure proper access. + */ + clear_bit(HT_AGG_STATE_OPERATIONAL, &tid_tx->state); - *state = HT_AGG_STATE_REQ_STOP_BA_MSK | - (initiator << HT_AGG_STATE_INITIATOR_SHIFT); + tid_tx->stop_initiator = initiator; ret = drv_ampdu_action(local, sta->sdata, IEEE80211_AMPDU_TX_STOP, @@ -174,15 +191,13 @@ static void sta_addba_resp_timer_expired(unsigned long data) u16 tid = *(u8 *)data; struct sta_info *sta = container_of((void *)data, struct sta_info, timer_to_tid[tid]); - u8 *state; - - state = &sta->ampdu_mlme.tid_state_tx[tid]; + struct tid_ampdu_tx *tid_tx; /* check if the TID waits for addBA response */ spin_lock_bh(&sta->lock); - if ((*state & (HT_ADDBA_REQUESTED_MSK | HT_ADDBA_RECEIVED_MSK | - HT_AGG_STATE_REQ_STOP_BA_MSK)) != - HT_ADDBA_REQUESTED_MSK) { + tid_tx = sta->ampdu_mlme.tid_tx[tid]; + if (!tid_tx || + test_bit(HT_AGG_STATE_RESPONSE_RECEIVED, &tid_tx->state)) { spin_unlock_bh(&sta->lock); #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "timer expired on tid %d but we are not " @@ -210,7 +225,7 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) struct sta_info *sta = container_of(pubsta, struct sta_info, sta); struct ieee80211_sub_if_data *sdata = sta->sdata; struct ieee80211_local *local = sdata->local; - u8 *state; + struct tid_ampdu_tx *tid_tx; int ret = 0; u16 start_seq_num; @@ -256,9 +271,9 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) goto err_unlock_sta; } - state = &sta->ampdu_mlme.tid_state_tx[tid]; + tid_tx = sta->ampdu_mlme.tid_tx[tid]; /* check if the TID is not in aggregation flow already */ - if (*state != HT_AGG_STATE_IDLE) { + if (tid_tx) { #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "BA request denied - session is not " "idle on tid %u\n", tid); @@ -279,9 +294,8 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) IEEE80211_QUEUE_STOP_REASON_AGGREGATION); /* prepare A-MPDU MLME for Tx aggregation */ - sta->ampdu_mlme.tid_tx[tid] = - kmalloc(sizeof(struct tid_ampdu_tx), GFP_ATOMIC); - if (!sta->ampdu_mlme.tid_tx[tid]) { + tid_tx = kzalloc(sizeof(struct tid_ampdu_tx), GFP_ATOMIC); + if (!tid_tx) { #ifdef CONFIG_MAC80211_HT_DEBUG if (net_ratelimit()) printk(KERN_ERR "allocate tx mlme to tid %d failed\n", @@ -291,33 +305,27 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) goto err_wake_queue; } - skb_queue_head_init(&sta->ampdu_mlme.tid_tx[tid]->pending); + skb_queue_head_init(&tid_tx->pending); /* Tx timer */ - sta->ampdu_mlme.tid_tx[tid]->addba_resp_timer.function = - sta_addba_resp_timer_expired; - sta->ampdu_mlme.tid_tx[tid]->addba_resp_timer.data = - (unsigned long)&sta->timer_to_tid[tid]; - init_timer(&sta->ampdu_mlme.tid_tx[tid]->addba_resp_timer); - - /* Ok, the Addba frame hasn't been sent yet, but if the driver calls the - * call back right away, it must see that the flow has begun */ - *state |= HT_ADDBA_REQUESTED_MSK; + tid_tx->addba_resp_timer.function = sta_addba_resp_timer_expired; + tid_tx->addba_resp_timer.data = (unsigned long)&sta->timer_to_tid[tid]; + init_timer(&tid_tx->addba_resp_timer); start_seq_num = sta->tid_seq[tid] >> 4; ret = drv_ampdu_action(local, sdata, IEEE80211_AMPDU_TX_START, pubsta, tid, &start_seq_num); - if (ret) { #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "BA request denied - HW unavailable for" " tid %d\n", tid); #endif /* CONFIG_MAC80211_HT_DEBUG */ - *state = HT_AGG_STATE_IDLE; goto err_free; } + rcu_assign_pointer(sta->ampdu_mlme.tid_tx[tid], tid_tx); + /* Driver vetoed or OKed, but we can take packets again now */ ieee80211_wake_queue_by_reason( &local->hw, ieee80211_ac_from_tid(tid), @@ -325,32 +333,30 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) spin_unlock(&local->ampdu_lock); + /* activate the timer for the recipient's addBA response */ + tid_tx->addba_resp_timer.expires = jiffies + ADDBA_RESP_INTERVAL; + add_timer(&tid_tx->addba_resp_timer); +#ifdef CONFIG_MAC80211_HT_DEBUG + printk(KERN_DEBUG "activated addBA response timer on tid %d\n", tid); +#endif + /* prepare tid data */ sta->ampdu_mlme.dialog_token_allocator++; - sta->ampdu_mlme.tid_tx[tid]->dialog_token = - sta->ampdu_mlme.dialog_token_allocator; - sta->ampdu_mlme.tid_tx[tid]->ssn = start_seq_num; + tid_tx->dialog_token = sta->ampdu_mlme.dialog_token_allocator; + tid_tx->ssn = start_seq_num; + + sta->ampdu_mlme.addba_req_num[tid]++; spin_unlock_bh(&sta->lock); /* send AddBA request */ ieee80211_send_addba_request(sdata, pubsta->addr, tid, - sta->ampdu_mlme.tid_tx[tid]->dialog_token, - sta->ampdu_mlme.tid_tx[tid]->ssn, + tid_tx->dialog_token, tid_tx->ssn, 0x40, 5000); - sta->ampdu_mlme.addba_req_num[tid]++; - /* activate the timer for the recipient's addBA response */ - sta->ampdu_mlme.tid_tx[tid]->addba_resp_timer.expires = - jiffies + ADDBA_RESP_INTERVAL; - add_timer(&sta->ampdu_mlme.tid_tx[tid]->addba_resp_timer); -#ifdef CONFIG_MAC80211_HT_DEBUG - printk(KERN_DEBUG "activated addBA response timer on tid %d\n", tid); -#endif return 0; err_free: - kfree(sta->ampdu_mlme.tid_tx[tid]); - sta->ampdu_mlme.tid_tx[tid] = NULL; + kfree(tid_tx); err_wake_queue: ieee80211_wake_queue_by_reason( &local->hw, ieee80211_ac_from_tid(tid), @@ -368,7 +374,8 @@ EXPORT_SYMBOL(ieee80211_start_tx_ba_session); * local->ampdu_lock across both calls. */ static void ieee80211_agg_splice_packets(struct ieee80211_local *local, - struct sta_info *sta, u16 tid) + struct tid_ampdu_tx *tid_tx, + u16 tid) { unsigned long flags; u16 queue = ieee80211_ac_from_tid(tid); @@ -377,31 +384,23 @@ static void ieee80211_agg_splice_packets(struct ieee80211_local *local, &local->hw, queue, IEEE80211_QUEUE_STOP_REASON_AGGREGATION); - if (!(sta->ampdu_mlme.tid_state_tx[tid] & HT_ADDBA_REQUESTED_MSK)) - return; - - if (WARN(!sta->ampdu_mlme.tid_tx[tid], - "TID %d gone but expected when splicing aggregates from" - "the pending queue\n", tid)) + if (WARN(!tid_tx, "TID %d gone but expected when splicing aggregates" + " from the pending queue\n", tid)) return; - if (!skb_queue_empty(&sta->ampdu_mlme.tid_tx[tid]->pending)) { + if (!skb_queue_empty(&tid_tx->pending)) { spin_lock_irqsave(&local->queue_stop_reason_lock, flags); /* copy over remaining packets */ - skb_queue_splice_tail_init( - &sta->ampdu_mlme.tid_tx[tid]->pending, - &local->pending[queue]); + skb_queue_splice_tail_init(&tid_tx->pending, + &local->pending[queue]); spin_unlock_irqrestore(&local->queue_stop_reason_lock, flags); } } -static void ieee80211_agg_splice_finish(struct ieee80211_local *local, - struct sta_info *sta, u16 tid) +static void ieee80211_agg_splice_finish(struct ieee80211_local *local, u16 tid) { - u16 queue = ieee80211_ac_from_tid(tid); - ieee80211_wake_queue_by_reason( - &local->hw, queue, + &local->hw, ieee80211_ac_from_tid(tid), IEEE80211_QUEUE_STOP_REASON_AGGREGATION); } @@ -409,19 +408,21 @@ static void ieee80211_agg_splice_finish(struct ieee80211_local *local, static void ieee80211_agg_tx_operational(struct ieee80211_local *local, struct sta_info *sta, u16 tid) { + lockdep_assert_held(&sta->lock); + #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "Aggregation is on for tid %d\n", tid); #endif spin_lock(&local->ampdu_lock); - ieee80211_agg_splice_packets(local, sta, tid); + ieee80211_agg_splice_packets(local, sta->ampdu_mlme.tid_tx[tid], tid); /* - * NB: we rely on sta->lock being taken in the TX - * processing here when adding to the pending queue, - * otherwise we could only change the state of the - * session to OPERATIONAL _here_. + * Now mark as operational. This will be visible + * in the TX path, and lets it go lock-free in + * the common case. */ - ieee80211_agg_splice_finish(local, sta, tid); + set_bit(HT_AGG_STATE_OPERATIONAL, &sta->ampdu_mlme.tid_tx[tid]->state); + ieee80211_agg_splice_finish(local, tid); spin_unlock(&local->ampdu_lock); drv_ampdu_action(local, sta->sdata, @@ -434,7 +435,7 @@ void ieee80211_start_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u16 tid) struct ieee80211_sub_if_data *sdata = vif_to_sdata(vif); struct ieee80211_local *local = sdata->local; struct sta_info *sta; - u8 *state; + struct tid_ampdu_tx *tid_tx; trace_api_start_tx_ba_cb(sdata, ra, tid); @@ -456,25 +457,22 @@ void ieee80211_start_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u16 tid) return; } - state = &sta->ampdu_mlme.tid_state_tx[tid]; spin_lock_bh(&sta->lock); + tid_tx = sta->ampdu_mlme.tid_tx[tid]; - if (WARN_ON(!(*state & HT_ADDBA_REQUESTED_MSK))) { + if (WARN_ON(!tid_tx)) { #ifdef CONFIG_MAC80211_HT_DEBUG - printk(KERN_DEBUG "addBA was not requested yet, state is %d\n", - *state); + printk(KERN_DEBUG "addBA was not requested!\n"); #endif spin_unlock_bh(&sta->lock); rcu_read_unlock(); return; } - if (WARN_ON(*state & HT_ADDBA_DRV_READY_MSK)) + if (WARN_ON(test_and_set_bit(HT_AGG_STATE_DRV_READY, &tid_tx->state))) goto out; - *state |= HT_ADDBA_DRV_READY_MSK; - - if (*state == HT_AGG_STATE_OPERATIONAL) + if (test_bit(HT_AGG_STATE_RESPONSE_RECEIVED, &tid_tx->state)) ieee80211_agg_tx_operational(local, sta, tid); out: @@ -512,14 +510,14 @@ EXPORT_SYMBOL(ieee80211_start_tx_ba_cb_irqsafe); int __ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, enum ieee80211_back_parties initiator) { - u8 *state; + struct tid_ampdu_tx *tid_tx; int ret; - /* check if the TID is in aggregation */ - state = &sta->ampdu_mlme.tid_state_tx[tid]; spin_lock_bh(&sta->lock); + tid_tx = sta->ampdu_mlme.tid_tx[tid]; - if (*state != HT_AGG_STATE_OPERATIONAL) { + /* check if the TID is in aggregation */ + if (!tid_tx || !test_bit(HT_AGG_STATE_OPERATIONAL, &tid_tx->state)) { ret = -ENOENT; goto unlock; } @@ -554,7 +552,7 @@ void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid) struct ieee80211_sub_if_data *sdata = vif_to_sdata(vif); struct ieee80211_local *local = sdata->local; struct sta_info *sta; - u8 *state; + struct tid_ampdu_tx *tid_tx; trace_api_stop_tx_ba_cb(sdata, ra, tid); @@ -580,39 +578,45 @@ void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid) rcu_read_unlock(); return; } - state = &sta->ampdu_mlme.tid_state_tx[tid]; - /* NOTE: no need to use sta->lock in this state check, as - * ieee80211_stop_tx_ba_session will let only one stop call to - * pass through per sta/tid - */ - if ((*state & HT_AGG_STATE_REQ_STOP_BA_MSK) == 0) { + spin_lock_bh(&sta->lock); + tid_tx = sta->ampdu_mlme.tid_tx[tid]; + + if (!tid_tx || !test_bit(HT_AGG_STATE_STOPPING, &tid_tx->state)) { #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "unexpected callback to A-MPDU stop\n"); #endif + spin_unlock_bh(&sta->lock); rcu_read_unlock(); return; } - if (*state & HT_AGG_STATE_INITIATOR_MSK) + if (tid_tx->stop_initiator == WLAN_BACK_INITIATOR) ieee80211_send_delba(sta->sdata, ra, tid, WLAN_BACK_INITIATOR, WLAN_REASON_QSTA_NOT_USE); - spin_lock_bh(&sta->lock); - spin_lock(&local->ampdu_lock); + /* + * When we get here, the TX path will not be lockless any more wrt. + * aggregation, since the OPERATIONAL bit has long been cleared. + * Thus it will block on getting the lock, if it occurs. So if we + * stop the queue now, we will not get any more packets, and any + * that might be being processed will wait for us here, thereby + * guaranteeing that no packets go to the tid_tx pending queue any + * more. + */ - ieee80211_agg_splice_packets(local, sta, tid); + spin_lock(&local->ampdu_lock); + ieee80211_agg_splice_packets(local, tid_tx, tid); - *state = HT_AGG_STATE_IDLE; - /* from now on packets are no longer put onto sta->pending */ - kfree(sta->ampdu_mlme.tid_tx[tid]); - sta->ampdu_mlme.tid_tx[tid] = NULL; + /* future packets must not find the tid_tx struct any more */ + rcu_assign_pointer(sta->ampdu_mlme.tid_tx[tid], NULL); - ieee80211_agg_splice_finish(local, sta, tid); + ieee80211_agg_splice_finish(local, tid); + call_rcu(&tid_tx->rcu_head, kfree_tid_tx); spin_unlock(&local->ampdu_lock); - spin_unlock_bh(&sta->lock); + spin_unlock_bh(&sta->lock); rcu_read_unlock(); } EXPORT_SYMBOL(ieee80211_stop_tx_ba_cb); @@ -649,40 +653,41 @@ void ieee80211_process_addba_resp(struct ieee80211_local *local, struct ieee80211_mgmt *mgmt, size_t len) { + struct tid_ampdu_tx *tid_tx; u16 capab, tid; - u8 *state; capab = le16_to_cpu(mgmt->u.action.u.addba_resp.capab); tid = (capab & IEEE80211_ADDBA_PARAM_TID_MASK) >> 2; - state = &sta->ampdu_mlme.tid_state_tx[tid]; - spin_lock_bh(&sta->lock); - if (!(*state & HT_ADDBA_REQUESTED_MSK)) + tid_tx = sta->ampdu_mlme.tid_tx[tid]; + + if (!tid_tx) goto out; - if (mgmt->u.action.u.addba_resp.dialog_token != - sta->ampdu_mlme.tid_tx[tid]->dialog_token) { + if (mgmt->u.action.u.addba_resp.dialog_token != tid_tx->dialog_token) { #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "wrong addBA response token, tid %d\n", tid); -#endif /* CONFIG_MAC80211_HT_DEBUG */ +#endif goto out; } - del_timer(&sta->ampdu_mlme.tid_tx[tid]->addba_resp_timer); + del_timer(&tid_tx->addba_resp_timer); #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "switched off addBA timer for tid %d\n", tid); -#endif /* CONFIG_MAC80211_HT_DEBUG */ +#endif if (le16_to_cpu(mgmt->u.action.u.addba_resp.status) == WLAN_STATUS_SUCCESS) { - u8 curstate = *state; - - *state |= HT_ADDBA_RECEIVED_MSK; + if (test_and_set_bit(HT_AGG_STATE_RESPONSE_RECEIVED, + &tid_tx->state)) { + /* ignore duplicate response */ + goto out; + } - if (*state != curstate && *state == HT_AGG_STATE_OPERATIONAL) + if (test_bit(HT_AGG_STATE_DRV_READY, &tid_tx->state)) ieee80211_agg_tx_operational(local, sta, tid); sta->ampdu_mlme.addba_req_num[tid] = 0; diff --git a/net/mac80211/debugfs_sta.c b/net/mac80211/debugfs_sta.c index 95c1ea4..8a74ffb 100644 --- a/net/mac80211/debugfs_sta.c +++ b/net/mac80211/debugfs_sta.c @@ -134,15 +134,15 @@ static ssize_t sta_agg_status_read(struct file *file, char __user *userbuf, sta->ampdu_mlme.tid_rx[i]->ssn : 0); p += scnprintf(p, sizeof(buf) + buf - p, "\t\t%x", - sta->ampdu_mlme.tid_state_tx[i]); + !!sta->ampdu_mlme.tid_tx[i]); p += scnprintf(p, sizeof(buf) + buf - p, "\t%#.2x", - sta->ampdu_mlme.tid_state_tx[i] ? + sta->ampdu_mlme.tid_tx[i] ? sta->ampdu_mlme.tid_tx[i]->dialog_token : 0); p += scnprintf(p, sizeof(buf) + buf - p, "\t%#.3x", - sta->ampdu_mlme.tid_state_tx[i] ? + sta->ampdu_mlme.tid_tx[i] ? sta->ampdu_mlme.tid_tx[i]->ssn : 0); p += scnprintf(p, sizeof(buf) + buf - p, "\t%03d", - sta->ampdu_mlme.tid_state_tx[i] ? + sta->ampdu_mlme.tid_tx[i] ? skb_queue_len(&sta->ampdu_mlme.tid_tx[i]->pending) : 0); p += scnprintf(p, sizeof(buf) + buf - p, "\n"); } diff --git a/net/mac80211/ht.c b/net/mac80211/ht.c index 2ab106a..1af173e 100644 --- a/net/mac80211/ht.c +++ b/net/mac80211/ht.c @@ -176,13 +176,8 @@ void ieee80211_process_delba(struct ieee80211_sub_if_data *sdata, if (initiator == WLAN_BACK_INITIATOR) __ieee80211_stop_rx_ba_session(sta, tid, WLAN_BACK_INITIATOR, 0); - else { /* WLAN_BACK_RECIPIENT */ - spin_lock_bh(&sta->lock); - if (sta->ampdu_mlme.tid_state_tx[tid] & HT_ADDBA_REQUESTED_MSK) - ___ieee80211_stop_tx_ba_session(sta, tid, - WLAN_BACK_RECIPIENT); - spin_unlock_bh(&sta->lock); - } + else + __ieee80211_stop_tx_ba_session(sta, tid, WLAN_BACK_RECIPIENT); } int ieee80211_send_smps_action(struct ieee80211_sub_if_data *sdata, diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index bafe610..71bdd8b 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -1119,8 +1119,6 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, int __ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, enum ieee80211_back_parties initiator); -int ___ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, - enum ieee80211_back_parties initiator); /* Spectrum management */ void ieee80211_process_measurement_req(struct ieee80211_sub_if_data *sdata, diff --git a/net/mac80211/rc80211_minstrel_ht.c b/net/mac80211/rc80211_minstrel_ht.c index c23f082..7a04951 100644 --- a/net/mac80211/rc80211_minstrel_ht.c +++ b/net/mac80211/rc80211_minstrel_ht.c @@ -365,7 +365,7 @@ minstrel_aggr_check(struct minstrel_priv *mp, struct ieee80211_sta *pubsta, stru return; tid = *ieee80211_get_qos_ctl(hdr) & IEEE80211_QOS_CTL_TID_MASK; - if (likely(sta->ampdu_mlme.tid_state_tx[tid] != HT_AGG_STATE_IDLE)) + if (likely(sta->ampdu_mlme.tid_tx[tid])) return; ieee80211_start_tx_ba_session(pubsta, tid); diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c index c426c57..06d8e00 100644 --- a/net/mac80211/sta_info.c +++ b/net/mac80211/sta_info.c @@ -246,14 +246,12 @@ struct sta_info *sta_info_alloc(struct ieee80211_sub_if_data *sdata, } for (i = 0; i < STA_TID_NUM; i++) { - /* timer_to_tid must be initialized with identity mapping to - * enable session_timer's data differentiation. refer to - * sta_rx_agg_session_timer_expired for useage */ + /* + * timer_to_tid must be initialized with identity mapping + * to enable session_timer's data differentiation. See + * sta_rx_agg_session_timer_expired for usage. + */ sta->timer_to_tid[i] = i; - /* tx */ - sta->ampdu_mlme.tid_state_tx[i] = HT_AGG_STATE_IDLE; - sta->ampdu_mlme.tid_tx[i] = NULL; - sta->ampdu_mlme.addba_req_num[i] = 0; } skb_queue_head_init(&sta->ps_tx_buf); skb_queue_head_init(&sta->tx_filtered); diff --git a/net/mac80211/sta_info.h b/net/mac80211/sta_info.h index 16864a6..05278e62 100644 --- a/net/mac80211/sta_info.h +++ b/net/mac80211/sta_info.h @@ -61,33 +61,40 @@ enum ieee80211_sta_info_flags { #define STA_TID_NUM 16 #define ADDBA_RESP_INTERVAL HZ -#define HT_AGG_MAX_RETRIES (0x3) +#define HT_AGG_MAX_RETRIES 0x3 -#define HT_AGG_STATE_INITIATOR_SHIFT (4) - -#define HT_ADDBA_REQUESTED_MSK BIT(0) -#define HT_ADDBA_DRV_READY_MSK BIT(1) -#define HT_ADDBA_RECEIVED_MSK BIT(2) -#define HT_AGG_STATE_REQ_STOP_BA_MSK BIT(3) -#define HT_AGG_STATE_INITIATOR_MSK BIT(HT_AGG_STATE_INITIATOR_SHIFT) -#define HT_AGG_STATE_IDLE (0x0) -#define HT_AGG_STATE_OPERATIONAL (HT_ADDBA_REQUESTED_MSK | \ - HT_ADDBA_DRV_READY_MSK | \ - HT_ADDBA_RECEIVED_MSK) +#define HT_AGG_STATE_DRV_READY 0 +#define HT_AGG_STATE_RESPONSE_RECEIVED 1 +#define HT_AGG_STATE_OPERATIONAL 2 +#define HT_AGG_STATE_STOPPING 3 /** * struct tid_ampdu_tx - TID aggregation information (Tx). * + * @rcu_head: rcu head for freeing structure * @addba_resp_timer: timer for peer's response to addba request * @pending: pending frames queue -- use sta's spinlock to protect * @ssn: Starting Sequence Number expected to be aggregated. * @dialog_token: dialog token for aggregation session + * @state: session state (see above) + * @stop_initiator: initiator of a session stop + * + * This structure is protected by RCU and the per-station + * spinlock. Assignments to the array holding it must hold + * the spinlock, only the TX path can access it under RCU + * lock-free if, and only if, the state has the flag + * %HT_AGG_STATE_OPERATIONAL set. Otherwise, the TX path + * must also acquire the spinlock and re-check the state, + * see comments in the tx code touching it. */ struct tid_ampdu_tx { + struct rcu_head rcu_head; struct timer_list addba_resp_timer; struct sk_buff_head pending; + unsigned long state; u16 ssn; u8 dialog_token; + u8 stop_initiator; }; /** @@ -129,7 +136,6 @@ struct tid_ampdu_rx { * struct sta_ampdu_mlme - STA aggregation information. * * @tid_rx: aggregation info for Rx per TID -- RCU protected - * @tid_state_tx: TID's state in Tx session state machine. * @tid_tx: aggregation info for Tx per TID * @addba_req_num: number of times addBA request has been sent. * @dialog_token_allocator: dialog token enumerator for each new session; @@ -138,7 +144,6 @@ struct sta_ampdu_mlme { /* rx */ struct tid_ampdu_rx *tid_rx[STA_TID_NUM]; /* tx */ - u8 tid_state_tx[STA_TID_NUM]; struct tid_ampdu_tx *tid_tx[STA_TID_NUM]; u8 addba_req_num[STA_TID_NUM]; u8 dialog_token_allocator; diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index 680bcb7..7bf1f9c 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -1092,6 +1092,54 @@ static bool __ieee80211_parse_tx_radiotap(struct ieee80211_tx_data *tx, return true; } +static bool ieee80211_tx_prep_agg(struct ieee80211_tx_data *tx, + struct sk_buff *skb, + struct ieee80211_tx_info *info, + struct tid_ampdu_tx *tid_tx, + int tid) +{ + bool queued = false; + + if (test_bit(HT_AGG_STATE_OPERATIONAL, &tid_tx->state)) { + info->flags |= IEEE80211_TX_CTL_AMPDU; + } else { + spin_lock(&tx->sta->lock); + /* + * Need to re-check now, because we may get here + * + * 1) in the window during which the setup is actually + * already done, but not marked yet because not all + * packets are spliced over to the driver pending + * queue yet -- if this happened we acquire the lock + * either before or after the splice happens, but + * need to recheck which of these cases happened. + * + * 2) during session teardown, if the OPERATIONAL bit + * was cleared due to the teardown but the pointer + * hasn't been assigned NULL yet (or we loaded it + * before it was assigned) -- in this case it may + * now be NULL which means we should just let the + * packet pass through because splicing the frames + * back is already done. + */ + tid_tx = tx->sta->ampdu_mlme.tid_tx[tid]; + + if (!tid_tx) { + /* do nothing, let packet pass through */ + } else if (test_bit(HT_AGG_STATE_OPERATIONAL, &tid_tx->state)) { + info->flags |= IEEE80211_TX_CTL_AMPDU; + } else { + queued = true; + info->control.vif = &tx->sdata->vif; + info->flags |= IEEE80211_TX_INTFL_NEED_TXPROCESSING; + __skb_queue_tail(&tid_tx->pending, skb); + } + spin_unlock(&tx->sta->lock); + } + + return queued; +} + /* * initialises @tx */ @@ -1104,8 +1152,7 @@ ieee80211_tx_prepare(struct ieee80211_sub_if_data *sdata, struct ieee80211_hdr *hdr; struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); int hdrlen, tid; - u8 *qc, *state; - bool queued = false; + u8 *qc; memset(tx, 0, sizeof(*tx)); tx->skb = skb; @@ -1157,35 +1204,16 @@ ieee80211_tx_prepare(struct ieee80211_sub_if_data *sdata, qc = ieee80211_get_qos_ctl(hdr); tid = *qc & IEEE80211_QOS_CTL_TID_MASK; - spin_lock(&tx->sta->lock); - /* - * XXX: This spinlock could be fairly expensive, but see the - * comment in agg-tx.c:ieee80211_agg_tx_operational(). - * One way to solve this would be to do something RCU-like - * for managing the tid_tx struct and using atomic bitops - * for the actual state -- by introducing an actual - * 'operational' bit that would be possible. It would - * require changing ieee80211_agg_tx_operational() to - * set that bit, and changing the way tid_tx is managed - * everywhere, including races between that bit and - * tid_tx going away (tid_tx being added can be easily - * committed to memory before the 'operational' bit). - */ - tid_tx = tx->sta->ampdu_mlme.tid_tx[tid]; - state = &tx->sta->ampdu_mlme.tid_state_tx[tid]; - if (*state == HT_AGG_STATE_OPERATIONAL) { - info->flags |= IEEE80211_TX_CTL_AMPDU; - } else if (*state != HT_AGG_STATE_IDLE) { - /* in progress */ - queued = true; - info->control.vif = &sdata->vif; - info->flags |= IEEE80211_TX_INTFL_NEED_TXPROCESSING; - __skb_queue_tail(&tid_tx->pending, skb); - } - spin_unlock(&tx->sta->lock); + tid_tx = rcu_dereference(tx->sta->ampdu_mlme.tid_tx[tid]); + if (tid_tx) { + bool queued; - if (unlikely(queued)) - return TX_QUEUED; + queued = ieee80211_tx_prep_agg(tx, skb, info, + tid_tx, tid); + + if (unlikely(queued)) + return TX_QUEUED; + } } if (is_multicast_ether_addr(hdr->addr1)) { -- cgit v1.1 From 5d22c89b9bea17a0e48e7534a9b237885e2c0809 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:40 +0200 Subject: mac80211: remove non-irqsafe aggregation callbacks The non-irqsafe aggregation start/stop done callbacks are currently only used by ath9k_htc, and can cause callbacks into the driver again. This might lead to locking issues, which will only get worse as we modify locking. To avoid trouble, remove the non-irqsafe versions and change ath9k_htc to use those instead. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 2 -- net/mac80211/ieee80211_i.h | 2 ++ 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index 7d8656d..5a7ef51 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -479,7 +479,6 @@ void ieee80211_start_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u16 tid) spin_unlock_bh(&sta->lock); rcu_read_unlock(); } -EXPORT_SYMBOL(ieee80211_start_tx_ba_cb); void ieee80211_start_tx_ba_cb_irqsafe(struct ieee80211_vif *vif, const u8 *ra, u16 tid) @@ -619,7 +618,6 @@ void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid) spin_unlock_bh(&sta->lock); rcu_read_unlock(); } -EXPORT_SYMBOL(ieee80211_stop_tx_ba_cb); void ieee80211_stop_tx_ba_cb_irqsafe(struct ieee80211_vif *vif, const u8 *ra, u16 tid) diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 71bdd8b..a3ae513 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -1119,6 +1119,8 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, int __ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, enum ieee80211_back_parties initiator); +void ieee80211_start_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u16 tid); +void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid); /* Spectrum management */ void ieee80211_process_measurement_req(struct ieee80211_sub_if_data *sdata, -- cgit v1.1 From a6a67db2bc89d2b1ff07e0817f11235c20d2c329 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:41 +0200 Subject: mac80211: refcount aggregation queue stop mac80211 currently maintains the ampdu_lock to avoid starting a queue due to one aggregation session while another aggregation session needs the queue stopped. We can do better, however, and instead refcount the queue stops for this particular purpose, thus removing the need for the lock. This will help making ampdu_action able to sleep. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 77 ++++++++++++++++++++++++++++------------------ net/mac80211/ieee80211_i.h | 8 +---- net/mac80211/main.c | 6 ++-- 3 files changed, 51 insertions(+), 40 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index 5a7ef51..7f2042d 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -220,6 +220,41 @@ static inline int ieee80211_ac_from_tid(int tid) return ieee802_1d_to_ac[tid & 7]; } +/* + * When multiple aggregation sessions on multiple stations + * are being created/destroyed simultaneously, we need to + * refcount the global queue stop caused by that in order + * to not get into a situation where one of the aggregation + * setup or teardown re-enables queues before the other is + * ready to handle that. + * + * These two functions take care of this issue by keeping + * a global "agg_queue_stop" refcount. + */ +static void __acquires(agg_queue) +ieee80211_stop_queue_agg(struct ieee80211_local *local, int tid) +{ + int queue = ieee80211_ac_from_tid(tid); + + if (atomic_inc_return(&local->agg_queue_stop[queue]) == 1) + ieee80211_stop_queue_by_reason( + &local->hw, queue, + IEEE80211_QUEUE_STOP_REASON_AGGREGATION); + __acquire(agg_queue); +} + +static void __releases(agg_queue) +ieee80211_wake_queue_agg(struct ieee80211_local *local, int tid) +{ + int queue = ieee80211_ac_from_tid(tid); + + if (atomic_dec_return(&local->agg_queue_stop[queue]) == 0) + ieee80211_wake_queue_by_reason( + &local->hw, queue, + IEEE80211_QUEUE_STOP_REASON_AGGREGATION); + __release(agg_queue); +} + int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) { struct sta_info *sta = container_of(pubsta, struct sta_info, sta); @@ -263,7 +298,6 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) } spin_lock_bh(&sta->lock); - spin_lock(&local->ampdu_lock); /* we have tried too many times, receiver does not want A-MPDU */ if (sta->ampdu_mlme.addba_req_num[tid] > HT_AGG_MAX_RETRIES) { @@ -289,9 +323,7 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) * which would require us to put them to the AC pending * afterwards which just makes the code more complex. */ - ieee80211_stop_queue_by_reason( - &local->hw, ieee80211_ac_from_tid(tid), - IEEE80211_QUEUE_STOP_REASON_AGGREGATION); + ieee80211_stop_queue_agg(local, tid); /* prepare A-MPDU MLME for Tx aggregation */ tid_tx = kzalloc(sizeof(struct tid_ampdu_tx), GFP_ATOMIC); @@ -327,11 +359,7 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) rcu_assign_pointer(sta->ampdu_mlme.tid_tx[tid], tid_tx); /* Driver vetoed or OKed, but we can take packets again now */ - ieee80211_wake_queue_by_reason( - &local->hw, ieee80211_ac_from_tid(tid), - IEEE80211_QUEUE_STOP_REASON_AGGREGATION); - - spin_unlock(&local->ampdu_lock); + ieee80211_wake_queue_agg(local, tid); /* activate the timer for the recipient's addBA response */ tid_tx->addba_resp_timer.expires = jiffies + ADDBA_RESP_INTERVAL; @@ -358,11 +386,8 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) err_free: kfree(tid_tx); err_wake_queue: - ieee80211_wake_queue_by_reason( - &local->hw, ieee80211_ac_from_tid(tid), - IEEE80211_QUEUE_STOP_REASON_AGGREGATION); + ieee80211_wake_queue_agg(local, tid); err_unlock_sta: - spin_unlock(&local->ampdu_lock); spin_unlock_bh(&sta->lock); return ret; } @@ -370,19 +395,16 @@ EXPORT_SYMBOL(ieee80211_start_tx_ba_session); /* * splice packets from the STA's pending to the local pending, - * requires a call to ieee80211_agg_splice_finish and holding - * local->ampdu_lock across both calls. + * requires a call to ieee80211_agg_splice_finish later */ -static void ieee80211_agg_splice_packets(struct ieee80211_local *local, - struct tid_ampdu_tx *tid_tx, - u16 tid) +static void __acquires(agg_queue) +ieee80211_agg_splice_packets(struct ieee80211_local *local, + struct tid_ampdu_tx *tid_tx, u16 tid) { + int queue = ieee80211_ac_from_tid(tid); unsigned long flags; - u16 queue = ieee80211_ac_from_tid(tid); - ieee80211_stop_queue_by_reason( - &local->hw, queue, - IEEE80211_QUEUE_STOP_REASON_AGGREGATION); + ieee80211_stop_queue_agg(local, tid); if (WARN(!tid_tx, "TID %d gone but expected when splicing aggregates" " from the pending queue\n", tid)) @@ -397,11 +419,10 @@ static void ieee80211_agg_splice_packets(struct ieee80211_local *local, } } -static void ieee80211_agg_splice_finish(struct ieee80211_local *local, u16 tid) +static void __releases(agg_queue) +ieee80211_agg_splice_finish(struct ieee80211_local *local, u16 tid) { - ieee80211_wake_queue_by_reason( - &local->hw, ieee80211_ac_from_tid(tid), - IEEE80211_QUEUE_STOP_REASON_AGGREGATION); + ieee80211_wake_queue_agg(local, tid); } /* caller must hold sta->lock */ @@ -414,7 +435,6 @@ static void ieee80211_agg_tx_operational(struct ieee80211_local *local, printk(KERN_DEBUG "Aggregation is on for tid %d\n", tid); #endif - spin_lock(&local->ampdu_lock); ieee80211_agg_splice_packets(local, sta->ampdu_mlme.tid_tx[tid], tid); /* * Now mark as operational. This will be visible @@ -423,7 +443,6 @@ static void ieee80211_agg_tx_operational(struct ieee80211_local *local, */ set_bit(HT_AGG_STATE_OPERATIONAL, &sta->ampdu_mlme.tid_tx[tid]->state); ieee80211_agg_splice_finish(local, tid); - spin_unlock(&local->ampdu_lock); drv_ampdu_action(local, sta->sdata, IEEE80211_AMPDU_TX_OPERATIONAL, @@ -604,7 +623,6 @@ void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid) * more. */ - spin_lock(&local->ampdu_lock); ieee80211_agg_splice_packets(local, tid_tx, tid); /* future packets must not find the tid_tx struct any more */ @@ -613,7 +631,6 @@ void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid) ieee80211_agg_splice_finish(local, tid); call_rcu(&tid_tx->rcu_head, kfree_tid_tx); - spin_unlock(&local->ampdu_lock); spin_unlock_bh(&sta->lock); rcu_read_unlock(); diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index a3ae513..8a91b5d 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -723,13 +723,7 @@ struct ieee80211_local { struct sk_buff_head pending[IEEE80211_MAX_QUEUES]; struct tasklet_struct tx_pending_tasklet; - /* - * This lock is used to prevent concurrent A-MPDU - * session start/stop processing, this thus also - * synchronises the ->ampdu_action() callback to - * drivers and limits it to one at a time. - */ - spinlock_t ampdu_lock; + atomic_t agg_queue_stop[IEEE80211_MAX_QUEUES]; /* number of interfaces with corresponding IFF_ flags */ atomic_t iff_allmultis, iff_promiscs; diff --git a/net/mac80211/main.c b/net/mac80211/main.c index a2d10d4..c2e46e8 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -463,8 +463,10 @@ struct ieee80211_hw *ieee80211_alloc_hw(size_t priv_data_len, sta_info_init(local); - for (i = 0; i < IEEE80211_MAX_QUEUES; i++) + for (i = 0; i < IEEE80211_MAX_QUEUES; i++) { skb_queue_head_init(&local->pending[i]); + atomic_set(&local->agg_queue_stop[i], 0); + } tasklet_init(&local->tx_pending_tasklet, ieee80211_tx_pending, (unsigned long)local); @@ -475,8 +477,6 @@ struct ieee80211_hw *ieee80211_alloc_hw(size_t priv_data_len, skb_queue_head_init(&local->skb_queue); skb_queue_head_init(&local->skb_queue_unreliable); - spin_lock_init(&local->ampdu_lock); - return local_to_hw(local); } EXPORT_SYMBOL(ieee80211_alloc_hw); -- cgit v1.1 From 0ab337032a0dfcd5f2527d3306d3deeba5f95b59 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:42 +0200 Subject: mac80211: make TX aggregation start/stop request async When the driver or rate control requests starting or stopping an aggregation session, that currently causes a direct callback into the driver, which could potentially cause locking problems. Also, the functions need to be callable from contexts that cannot sleep, and thus will interfere with making the ampdu_action callback sleeping. To address these issues, add a new work item for each station that will process any start or stop requests out of line. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 177 ++++++++++++++++++++++++++++++++------------- net/mac80211/debugfs_sta.c | 5 +- net/mac80211/ht.c | 2 + net/mac80211/ieee80211_i.h | 1 + net/mac80211/sta_info.c | 1 + net/mac80211/sta_info.h | 6 +- net/mac80211/tx.c | 5 ++ 7 files changed, 139 insertions(+), 58 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index 7f2042d..ee8bfe1 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -146,6 +146,13 @@ static int ___ieee80211_stop_tx_ba_session( if (WARN_ON(!tid_tx)) return -ENOENT; + if (test_bit(HT_AGG_STATE_WANT_START, &tid_tx->state)) { + /* not even started yet! */ + rcu_assign_pointer(sta->ampdu_mlme.tid_tx[tid], NULL); + call_rcu(&tid_tx->rcu_head, kfree_tid_tx); + return 0; + } + #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "Tx BA session stop requested for %pM tid %u\n", sta->sta.addr, tid); @@ -255,6 +262,94 @@ ieee80211_wake_queue_agg(struct ieee80211_local *local, int tid) __release(agg_queue); } +static void ieee80211_tx_ba_session_handle_start(struct sta_info *sta, int tid) +{ + struct tid_ampdu_tx *tid_tx = sta->ampdu_mlme.tid_tx[tid]; + struct ieee80211_local *local = sta->local; + struct ieee80211_sub_if_data *sdata = sta->sdata; + u16 start_seq_num; + int ret; + + /* + * While we're asking the driver about the aggregation, + * stop the AC queue so that we don't have to worry + * about frames that came in while we were doing that, + * which would require us to put them to the AC pending + * afterwards which just makes the code more complex. + */ + ieee80211_stop_queue_agg(local, tid); + + clear_bit(HT_AGG_STATE_WANT_START, &tid_tx->state); + + /* + * This might be off by one due to a race that we can't + * really prevent here without synchronize_net() which + * can't be called now. + */ + start_seq_num = sta->tid_seq[tid] >> 4; + + ret = drv_ampdu_action(local, sdata, IEEE80211_AMPDU_TX_START, + &sta->sta, tid, &start_seq_num); + if (ret) { +#ifdef CONFIG_MAC80211_HT_DEBUG + printk(KERN_DEBUG "BA request denied - HW unavailable for" + " tid %d\n", tid); +#endif + rcu_assign_pointer(sta->ampdu_mlme.tid_tx[tid], NULL); + ieee80211_wake_queue_agg(local, tid); + call_rcu(&tid_tx->rcu_head, kfree_tid_tx); + return; + } + + /* we can take packets again now */ + ieee80211_wake_queue_agg(local, tid); + + /* activate the timer for the recipient's addBA response */ + mod_timer(&tid_tx->addba_resp_timer, jiffies + ADDBA_RESP_INTERVAL); +#ifdef CONFIG_MAC80211_HT_DEBUG + printk(KERN_DEBUG "activated addBA response timer on tid %d\n", tid); +#endif + + sta->ampdu_mlme.addba_req_num[tid]++; + + /* send AddBA request */ + ieee80211_send_addba_request(sdata, sta->sta.addr, tid, + tid_tx->dialog_token, start_seq_num, + 0x40, 5000); +} + +void ieee80211_tx_ba_session_work(struct work_struct *work) +{ + struct sta_info *sta = + container_of(work, struct sta_info, ampdu_mlme.work); + struct tid_ampdu_tx *tid_tx; + int tid; + + /* + * When this flag is set, new sessions should be + * blocked, and existing sessions will be torn + * down by the code that set the flag, so this + * need not run. + */ + if (test_sta_flags(sta, WLAN_STA_BLOCK_BA)) + return; + + spin_lock_bh(&sta->lock); + for (tid = 0; tid < STA_TID_NUM; tid++) { + tid_tx = sta->ampdu_mlme.tid_tx[tid]; + if (!tid_tx) + continue; + + if (test_bit(HT_AGG_STATE_WANT_START, &tid_tx->state)) + ieee80211_tx_ba_session_handle_start(sta, tid); + else if (test_and_clear_bit(HT_AGG_STATE_WANT_STOP, + &tid_tx->state)) + ___ieee80211_stop_tx_ba_session(sta, tid, + WLAN_BACK_INITIATOR); + } + spin_unlock_bh(&sta->lock); +} + int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) { struct sta_info *sta = container_of(pubsta, struct sta_info, sta); @@ -262,7 +357,6 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) struct ieee80211_local *local = sdata->local; struct tid_ampdu_tx *tid_tx; int ret = 0; - u16 start_seq_num; trace_api_start_tx_ba_session(pubsta, tid); @@ -316,15 +410,6 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) goto err_unlock_sta; } - /* - * While we're asking the driver about the aggregation, - * stop the AC queue so that we don't have to worry - * about frames that came in while we were doing that, - * which would require us to put them to the AC pending - * afterwards which just makes the code more complex. - */ - ieee80211_stop_queue_agg(local, tid); - /* prepare A-MPDU MLME for Tx aggregation */ tid_tx = kzalloc(sizeof(struct tid_ampdu_tx), GFP_ATOMIC); if (!tid_tx) { @@ -334,59 +419,27 @@ int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) tid); #endif ret = -ENOMEM; - goto err_wake_queue; + goto err_unlock_sta; } skb_queue_head_init(&tid_tx->pending); + __set_bit(HT_AGG_STATE_WANT_START, &tid_tx->state); /* Tx timer */ tid_tx->addba_resp_timer.function = sta_addba_resp_timer_expired; tid_tx->addba_resp_timer.data = (unsigned long)&sta->timer_to_tid[tid]; init_timer(&tid_tx->addba_resp_timer); - start_seq_num = sta->tid_seq[tid] >> 4; - - ret = drv_ampdu_action(local, sdata, IEEE80211_AMPDU_TX_START, - pubsta, tid, &start_seq_num); - if (ret) { -#ifdef CONFIG_MAC80211_HT_DEBUG - printk(KERN_DEBUG "BA request denied - HW unavailable for" - " tid %d\n", tid); -#endif /* CONFIG_MAC80211_HT_DEBUG */ - goto err_free; - } - - rcu_assign_pointer(sta->ampdu_mlme.tid_tx[tid], tid_tx); - - /* Driver vetoed or OKed, but we can take packets again now */ - ieee80211_wake_queue_agg(local, tid); - - /* activate the timer for the recipient's addBA response */ - tid_tx->addba_resp_timer.expires = jiffies + ADDBA_RESP_INTERVAL; - add_timer(&tid_tx->addba_resp_timer); -#ifdef CONFIG_MAC80211_HT_DEBUG - printk(KERN_DEBUG "activated addBA response timer on tid %d\n", tid); -#endif - - /* prepare tid data */ + /* assign a dialog token */ sta->ampdu_mlme.dialog_token_allocator++; tid_tx->dialog_token = sta->ampdu_mlme.dialog_token_allocator; - tid_tx->ssn = start_seq_num; - - sta->ampdu_mlme.addba_req_num[tid]++; - spin_unlock_bh(&sta->lock); + /* finally, assign it to the array */ + rcu_assign_pointer(sta->ampdu_mlme.tid_tx[tid], tid_tx); - /* send AddBA request */ - ieee80211_send_addba_request(sdata, pubsta->addr, tid, - tid_tx->dialog_token, tid_tx->ssn, - 0x40, 5000); - return 0; + ieee80211_queue_work(&local->hw, &sta->ampdu_mlme.work); - err_free: - kfree(tid_tx); - err_wake_queue: - ieee80211_wake_queue_agg(local, tid); + /* this flow continues off the work */ err_unlock_sta: spin_unlock_bh(&sta->lock); return ret; @@ -534,8 +587,7 @@ int __ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, spin_lock_bh(&sta->lock); tid_tx = sta->ampdu_mlme.tid_tx[tid]; - /* check if the TID is in aggregation */ - if (!tid_tx || !test_bit(HT_AGG_STATE_OPERATIONAL, &tid_tx->state)) { + if (!tid_tx) { ret = -ENOENT; goto unlock; } @@ -552,6 +604,8 @@ int ieee80211_stop_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) struct sta_info *sta = container_of(pubsta, struct sta_info, sta); struct ieee80211_sub_if_data *sdata = sta->sdata; struct ieee80211_local *local = sdata->local; + struct tid_ampdu_tx *tid_tx; + int ret = 0; trace_api_stop_tx_ba_session(pubsta, tid); @@ -561,7 +615,26 @@ int ieee80211_stop_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) if (tid >= STA_TID_NUM) return -EINVAL; - return __ieee80211_stop_tx_ba_session(sta, tid, WLAN_BACK_INITIATOR); + spin_lock_bh(&sta->lock); + tid_tx = sta->ampdu_mlme.tid_tx[tid]; + + if (!tid_tx) { + ret = -ENOENT; + goto unlock; + } + + if (test_bit(HT_AGG_STATE_STOPPING, &tid_tx->state)) { + /* already in progress stopping it */ + ret = 0; + goto unlock; + } + + set_bit(HT_AGG_STATE_WANT_STOP, &tid_tx->state); + ieee80211_queue_work(&local->hw, &sta->ampdu_mlme.work); + + unlock: + spin_unlock_bh(&sta->lock); + return ret; } EXPORT_SYMBOL(ieee80211_stop_tx_ba_session); diff --git a/net/mac80211/debugfs_sta.c b/net/mac80211/debugfs_sta.c index 8a74ffb..76839d4 100644 --- a/net/mac80211/debugfs_sta.c +++ b/net/mac80211/debugfs_sta.c @@ -121,7 +121,7 @@ static ssize_t sta_agg_status_read(struct file *file, char __user *userbuf, p += scnprintf(p, sizeof(buf) + buf - p, "next dialog_token: %#02x\n", sta->ampdu_mlme.dialog_token_allocator + 1); p += scnprintf(p, sizeof(buf) + buf - p, - "TID\t\tRX active\tDTKN\tSSN\t\tTX\tDTKN\tSSN\tpending\n"); + "TID\t\tRX active\tDTKN\tSSN\t\tTX\tDTKN\tpending\n"); for (i = 0; i < STA_TID_NUM; i++) { p += scnprintf(p, sizeof(buf) + buf - p, "%02d", i); p += scnprintf(p, sizeof(buf) + buf - p, "\t\t%x", @@ -138,9 +138,6 @@ static ssize_t sta_agg_status_read(struct file *file, char __user *userbuf, p += scnprintf(p, sizeof(buf) + buf - p, "\t%#.2x", sta->ampdu_mlme.tid_tx[i] ? sta->ampdu_mlme.tid_tx[i]->dialog_token : 0); - p += scnprintf(p, sizeof(buf) + buf - p, "\t%#.3x", - sta->ampdu_mlme.tid_tx[i] ? - sta->ampdu_mlme.tid_tx[i]->ssn : 0); p += scnprintf(p, sizeof(buf) + buf - p, "\t%03d", sta->ampdu_mlme.tid_tx[i] ? skb_queue_len(&sta->ampdu_mlme.tid_tx[i]->pending) : 0); diff --git a/net/mac80211/ht.c b/net/mac80211/ht.c index 1af173e..4dfba78 100644 --- a/net/mac80211/ht.c +++ b/net/mac80211/ht.c @@ -105,6 +105,8 @@ void ieee80211_sta_tear_down_BA_sessions(struct sta_info *sta) { int i; + cancel_work_sync(&sta->ampdu_mlme.work); + for (i = 0; i < STA_TID_NUM; i++) { __ieee80211_stop_tx_ba_session(sta, i, WLAN_BACK_INITIATOR); __ieee80211_stop_rx_ba_session(sta, i, WLAN_BACK_RECIPIENT, diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 8a91b5d..aec84c3 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -1115,6 +1115,7 @@ int __ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, enum ieee80211_back_parties initiator); void ieee80211_start_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u16 tid); void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid); +void ieee80211_tx_ba_session_work(struct work_struct *work); /* Spectrum management */ void ieee80211_process_measurement_req(struct ieee80211_sub_if_data *sdata, diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c index 06d8e00..8aa8558 100644 --- a/net/mac80211/sta_info.c +++ b/net/mac80211/sta_info.c @@ -235,6 +235,7 @@ struct sta_info *sta_info_alloc(struct ieee80211_sub_if_data *sdata, spin_lock_init(&sta->lock); spin_lock_init(&sta->flaglock); INIT_WORK(&sta->drv_unblock_wk, sta_unblock); + INIT_WORK(&sta->ampdu_mlme.work, ieee80211_tx_ba_session_work); memcpy(sta->sta.addr, addr, ETH_ALEN); sta->local = local; diff --git a/net/mac80211/sta_info.h b/net/mac80211/sta_info.h index 05278e62..040cbb0 100644 --- a/net/mac80211/sta_info.h +++ b/net/mac80211/sta_info.h @@ -67,6 +67,8 @@ enum ieee80211_sta_info_flags { #define HT_AGG_STATE_RESPONSE_RECEIVED 1 #define HT_AGG_STATE_OPERATIONAL 2 #define HT_AGG_STATE_STOPPING 3 +#define HT_AGG_STATE_WANT_START 4 +#define HT_AGG_STATE_WANT_STOP 5 /** * struct tid_ampdu_tx - TID aggregation information (Tx). @@ -74,7 +76,6 @@ enum ieee80211_sta_info_flags { * @rcu_head: rcu head for freeing structure * @addba_resp_timer: timer for peer's response to addba request * @pending: pending frames queue -- use sta's spinlock to protect - * @ssn: Starting Sequence Number expected to be aggregated. * @dialog_token: dialog token for aggregation session * @state: session state (see above) * @stop_initiator: initiator of a session stop @@ -92,7 +93,6 @@ struct tid_ampdu_tx { struct timer_list addba_resp_timer; struct sk_buff_head pending; unsigned long state; - u16 ssn; u8 dialog_token; u8 stop_initiator; }; @@ -139,11 +139,13 @@ struct tid_ampdu_rx { * @tid_tx: aggregation info for Tx per TID * @addba_req_num: number of times addBA request has been sent. * @dialog_token_allocator: dialog token enumerator for each new session; + * @work: work struct for starting/stopping aggregation */ struct sta_ampdu_mlme { /* rx */ struct tid_ampdu_rx *tid_rx[STA_TID_NUM]; /* tx */ + struct work_struct work; struct tid_ampdu_tx *tid_tx[STA_TID_NUM]; u8 addba_req_num[STA_TID_NUM]; u8 dialog_token_allocator; diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index 7bf1f9c..698d471 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -1102,6 +1102,11 @@ static bool ieee80211_tx_prep_agg(struct ieee80211_tx_data *tx, if (test_bit(HT_AGG_STATE_OPERATIONAL, &tid_tx->state)) { info->flags |= IEEE80211_TX_CTL_AMPDU; + } else if (test_bit(HT_AGG_STATE_WANT_START, &tid_tx->state)) { + /* + * nothing -- this aggregation session is being started + * but that might still fail with the driver + */ } else { spin_lock(&tx->sta->lock); /* -- cgit v1.1 From 67c282c00c9c06733aae229662d209957f6d23a7 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:43 +0200 Subject: mac80211: move BA session work Move the block-ack session works into common code, since it will be needed for RX agg too in the next patches. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 39 +++------------------------------------ net/mac80211/ht.c | 32 ++++++++++++++++++++++++++++++++ net/mac80211/ieee80211_i.h | 5 ++++- net/mac80211/sta_info.c | 2 +- 4 files changed, 40 insertions(+), 38 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index ee8bfe1..9a00a79 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -133,9 +133,8 @@ static void kfree_tid_tx(struct rcu_head *rcu_head) kfree(tid_tx); } -static int ___ieee80211_stop_tx_ba_session( - struct sta_info *sta, u16 tid, - enum ieee80211_back_parties initiator) +int ___ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, + enum ieee80211_back_parties initiator) { struct ieee80211_local *local = sta->local; struct tid_ampdu_tx *tid_tx = sta->ampdu_mlme.tid_tx[tid]; @@ -262,7 +261,7 @@ ieee80211_wake_queue_agg(struct ieee80211_local *local, int tid) __release(agg_queue); } -static void ieee80211_tx_ba_session_handle_start(struct sta_info *sta, int tid) +void ieee80211_tx_ba_session_handle_start(struct sta_info *sta, int tid) { struct tid_ampdu_tx *tid_tx = sta->ampdu_mlme.tid_tx[tid]; struct ieee80211_local *local = sta->local; @@ -318,38 +317,6 @@ static void ieee80211_tx_ba_session_handle_start(struct sta_info *sta, int tid) 0x40, 5000); } -void ieee80211_tx_ba_session_work(struct work_struct *work) -{ - struct sta_info *sta = - container_of(work, struct sta_info, ampdu_mlme.work); - struct tid_ampdu_tx *tid_tx; - int tid; - - /* - * When this flag is set, new sessions should be - * blocked, and existing sessions will be torn - * down by the code that set the flag, so this - * need not run. - */ - if (test_sta_flags(sta, WLAN_STA_BLOCK_BA)) - return; - - spin_lock_bh(&sta->lock); - for (tid = 0; tid < STA_TID_NUM; tid++) { - tid_tx = sta->ampdu_mlme.tid_tx[tid]; - if (!tid_tx) - continue; - - if (test_bit(HT_AGG_STATE_WANT_START, &tid_tx->state)) - ieee80211_tx_ba_session_handle_start(sta, tid); - else if (test_and_clear_bit(HT_AGG_STATE_WANT_STOP, - &tid_tx->state)) - ___ieee80211_stop_tx_ba_session(sta, tid, - WLAN_BACK_INITIATOR); - } - spin_unlock_bh(&sta->lock); -} - int ieee80211_start_tx_ba_session(struct ieee80211_sta *pubsta, u16 tid) { struct sta_info *sta = container_of(pubsta, struct sta_info, sta); diff --git a/net/mac80211/ht.c b/net/mac80211/ht.c index 4dfba78..531a19d 100644 --- a/net/mac80211/ht.c +++ b/net/mac80211/ht.c @@ -114,6 +114,38 @@ void ieee80211_sta_tear_down_BA_sessions(struct sta_info *sta) } } +void ieee80211_ba_session_work(struct work_struct *work) +{ + struct sta_info *sta = + container_of(work, struct sta_info, ampdu_mlme.work); + struct tid_ampdu_tx *tid_tx; + int tid; + + /* + * When this flag is set, new sessions should be + * blocked, and existing sessions will be torn + * down by the code that set the flag, so this + * need not run. + */ + if (test_sta_flags(sta, WLAN_STA_BLOCK_BA)) + return; + + spin_lock_bh(&sta->lock); + for (tid = 0; tid < STA_TID_NUM; tid++) { + tid_tx = sta->ampdu_mlme.tid_tx[tid]; + if (!tid_tx) + continue; + + if (test_bit(HT_AGG_STATE_WANT_START, &tid_tx->state)) + ieee80211_tx_ba_session_handle_start(sta, tid); + else if (test_and_clear_bit(HT_AGG_STATE_WANT_STOP, + &tid_tx->state)) + ___ieee80211_stop_tx_ba_session(sta, tid, + WLAN_BACK_INITIATOR); + } + spin_unlock_bh(&sta->lock); +} + void ieee80211_send_delba(struct ieee80211_sub_if_data *sdata, const u8 *da, u16 tid, u16 initiator, u16 reason_code) diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index aec84c3..628a53d 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -1113,9 +1113,12 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, int __ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, enum ieee80211_back_parties initiator); +int ___ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, + enum ieee80211_back_parties initiator); void ieee80211_start_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u16 tid); void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid); -void ieee80211_tx_ba_session_work(struct work_struct *work); +void ieee80211_ba_session_work(struct work_struct *work); +void ieee80211_tx_ba_session_handle_start(struct sta_info *sta, int tid); /* Spectrum management */ void ieee80211_process_measurement_req(struct ieee80211_sub_if_data *sdata, diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c index 8aa8558..bd726c6 100644 --- a/net/mac80211/sta_info.c +++ b/net/mac80211/sta_info.c @@ -235,7 +235,7 @@ struct sta_info *sta_info_alloc(struct ieee80211_sub_if_data *sdata, spin_lock_init(&sta->lock); spin_lock_init(&sta->flaglock); INIT_WORK(&sta->drv_unblock_wk, sta_unblock); - INIT_WORK(&sta->ampdu_mlme.work, ieee80211_tx_ba_session_work); + INIT_WORK(&sta->ampdu_mlme.work, ieee80211_ba_session_work); memcpy(sta->sta.addr, addr, ETH_ALEN); sta->local = local; -- cgit v1.1 From 7c3b1dd8f483bf4bcb9776cd9cd3b37dd327d7ea Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:44 +0200 Subject: mac80211: defer RX agg session teardown to work Since we want the code to be able to sleep in the future, it must not be called from the timer directly. To prepare, move it out into the aggregation work. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-rx.c | 24 ++++++++++-------------- net/mac80211/ht.c | 5 +++++ net/mac80211/ieee80211_i.h | 2 ++ net/mac80211/sta_info.h | 3 +++ 4 files changed, 20 insertions(+), 14 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-rx.c b/net/mac80211/agg-rx.c index bbf36d9..1c4320b 100644 --- a/net/mac80211/agg-rx.c +++ b/net/mac80211/agg-rx.c @@ -32,21 +32,18 @@ static void ieee80211_free_tid_rx(struct rcu_head *h) kfree(tid_rx); } -static void ___ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, - u16 initiator, u16 reason, - bool from_timer) +void ___ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, + u16 initiator, u16 reason) { struct ieee80211_local *local = sta->local; struct tid_ampdu_rx *tid_rx; - spin_lock_bh(&sta->lock); + lockdep_assert_held(&sta->lock); tid_rx = sta->ampdu_mlme.tid_rx[tid]; - if (!tid_rx) { - spin_unlock_bh(&sta->lock); + if (!tid_rx) return; - } rcu_assign_pointer(sta->ampdu_mlme.tid_rx[tid], NULL); @@ -65,10 +62,7 @@ static void ___ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, ieee80211_send_delba(sta->sdata, sta->sta.addr, tid, 0, reason); - spin_unlock_bh(&sta->lock); - - if (!from_timer) - del_timer_sync(&tid_rx->session_timer); + del_timer_sync(&tid_rx->session_timer); call_rcu(&tid_rx->rcu_head, ieee80211_free_tid_rx); } @@ -76,7 +70,9 @@ static void ___ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, void __ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, u16 initiator, u16 reason) { - ___ieee80211_stop_rx_ba_session(sta, tid, initiator, reason, false); + spin_lock_bh(&sta->lock); + ___ieee80211_stop_rx_ba_session(sta, tid, initiator, reason); + spin_unlock_bh(&sta->lock); } /* @@ -97,8 +93,8 @@ static void sta_rx_agg_session_timer_expired(unsigned long data) #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "rx session timer expired on tid %d\n", (u16)*ptid); #endif - ___ieee80211_stop_rx_ba_session(sta, *ptid, WLAN_BACK_RECIPIENT, - WLAN_REASON_QSTA_TIMEOUT, true); + set_bit(*ptid, sta->ampdu_mlme.tid_rx_timer_expired); + ieee80211_queue_work(&sta->local->hw, &sta->ampdu_mlme.work); } static void ieee80211_send_addba_resp(struct ieee80211_sub_if_data *sdata, u8 *da, u16 tid, diff --git a/net/mac80211/ht.c b/net/mac80211/ht.c index 531a19d..730f808 100644 --- a/net/mac80211/ht.c +++ b/net/mac80211/ht.c @@ -132,6 +132,11 @@ void ieee80211_ba_session_work(struct work_struct *work) spin_lock_bh(&sta->lock); for (tid = 0; tid < STA_TID_NUM; tid++) { + if (test_and_clear_bit(tid, sta->ampdu_mlme.tid_rx_timer_expired)) + ___ieee80211_stop_rx_ba_session( + sta, tid, WLAN_BACK_RECIPIENT, + WLAN_REASON_QSTA_TIMEOUT); + tid_tx = sta->ampdu_mlme.tid_tx[tid]; if (!tid_tx) continue; diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 628a53d..9d753a0 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -1096,6 +1096,8 @@ int ieee80211_send_smps_action(struct ieee80211_sub_if_data *sdata, enum ieee80211_smps_mode smps, const u8 *da, const u8 *bssid); +void ___ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, + u16 initiator, u16 reason); void __ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, u16 initiator, u16 reason); void ieee80211_sta_tear_down_BA_sessions(struct sta_info *sta); diff --git a/net/mac80211/sta_info.h b/net/mac80211/sta_info.h index 040cbb0..500bafe 100644 --- a/net/mac80211/sta_info.h +++ b/net/mac80211/sta_info.h @@ -140,10 +140,13 @@ struct tid_ampdu_rx { * @addba_req_num: number of times addBA request has been sent. * @dialog_token_allocator: dialog token enumerator for each new session; * @work: work struct for starting/stopping aggregation + * @tid_rx_timer_expired: bitmap indicating on which TIDs the + * RX timer expired until the work for it runs */ struct sta_ampdu_mlme { /* rx */ struct tid_ampdu_rx *tid_rx[STA_TID_NUM]; + unsigned long tid_rx_timer_expired[BITS_TO_LONGS(STA_TID_NUM)]; /* tx */ struct work_struct work; struct tid_ampdu_tx *tid_tx[STA_TID_NUM]; -- cgit v1.1 From f955ebb44798e0058c987a0817810d2542e87349 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:45 +0200 Subject: mac80211: fix RX aggregation timer I noticed that when there was _no_ traffic at all on a given aggregation session, it would never time out. This won't happen unless you forced creating a session, but fix it anyway. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-rx.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'net') diff --git a/net/mac80211/agg-rx.c b/net/mac80211/agg-rx.c index 1c4320b..aa858a0 100644 --- a/net/mac80211/agg-rx.c +++ b/net/mac80211/agg-rx.c @@ -274,6 +274,10 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, /* activate it for RX */ rcu_assign_pointer(sta->ampdu_mlme.tid_rx[tid], tid_agg_rx); + + if (timeout) + mod_timer(&tid_agg_rx->session_timer, TU_TO_EXP_TIME(timeout)); + end: spin_unlock_bh(&sta->lock); -- cgit v1.1 From a93e364430de7f7a5f4eedd604ad1ab3d825dde5 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:46 +0200 Subject: mac80211: change RX aggregation locking To prepare for allowing drivers to sleep in ampdu_action, change the locking in the RX aggregation code to use a mutex, so that it would already allow drivers to sleep. But explicitly disable BHs around the callback for now since the TX part cannot yet sleep, and drivers' locking might require it. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-rx.c | 12 ++++++------ net/mac80211/driver-ops.h | 2 ++ net/mac80211/ht.c | 6 +++++- net/mac80211/iface.c | 8 ++++---- net/mac80211/sta_info.c | 1 + net/mac80211/sta_info.h | 3 +++ 6 files changed, 21 insertions(+), 11 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-rx.c b/net/mac80211/agg-rx.c index aa858a0..a843df2 100644 --- a/net/mac80211/agg-rx.c +++ b/net/mac80211/agg-rx.c @@ -6,7 +6,7 @@ * Copyright 2005-2006, Devicescape Software, Inc. * Copyright 2006-2007 Jiri Benc * Copyright 2007, Michael Wu - * Copyright 2007-2008, Intel Corporation + * Copyright 2007-2010, Intel Corporation * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -38,7 +38,7 @@ void ___ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, struct ieee80211_local *local = sta->local; struct tid_ampdu_rx *tid_rx; - lockdep_assert_held(&sta->lock); + lockdep_assert_held(&sta->ampdu_mlme.mtx); tid_rx = sta->ampdu_mlme.tid_rx[tid]; @@ -70,9 +70,9 @@ void ___ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, void __ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, u16 initiator, u16 reason) { - spin_lock_bh(&sta->lock); + mutex_lock(&sta->ampdu_mlme.mtx); ___ieee80211_stop_rx_ba_session(sta, tid, initiator, reason); - spin_unlock_bh(&sta->lock); + mutex_unlock(&sta->ampdu_mlme.mtx); } /* @@ -205,7 +205,7 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, /* examine state machine */ - spin_lock_bh(&sta->lock); + mutex_lock(&sta->ampdu_mlme.mtx); if (sta->ampdu_mlme.tid_rx[tid]) { #ifdef CONFIG_MAC80211_HT_DEBUG @@ -279,7 +279,7 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, mod_timer(&tid_agg_rx->session_timer, TU_TO_EXP_TIME(timeout)); end: - spin_unlock_bh(&sta->lock); + mutex_unlock(&sta->ampdu_mlme.mtx); end_no_lock: ieee80211_send_addba_resp(sta->sdata, sta->sta.addr, tid, diff --git a/net/mac80211/driver-ops.h b/net/mac80211/driver-ops.h index 7d18a32..e5e7ef1 100644 --- a/net/mac80211/driver-ops.h +++ b/net/mac80211/driver-ops.h @@ -349,9 +349,11 @@ static inline int drv_ampdu_action(struct ieee80211_local *local, u16 *ssn) { int ret = -EOPNOTSUPP; + local_bh_disable(); if (local->ops->ampdu_action) ret = local->ops->ampdu_action(&local->hw, &sdata->vif, action, sta, tid, ssn); + local_bh_enable(); trace_drv_ampdu_action(local, sdata, action, sta, tid, ssn, ret); return ret; } diff --git a/net/mac80211/ht.c b/net/mac80211/ht.c index 730f808..e29be64 100644 --- a/net/mac80211/ht.c +++ b/net/mac80211/ht.c @@ -130,13 +130,17 @@ void ieee80211_ba_session_work(struct work_struct *work) if (test_sta_flags(sta, WLAN_STA_BLOCK_BA)) return; - spin_lock_bh(&sta->lock); + mutex_lock(&sta->ampdu_mlme.mtx); for (tid = 0; tid < STA_TID_NUM; tid++) { if (test_and_clear_bit(tid, sta->ampdu_mlme.tid_rx_timer_expired)) ___ieee80211_stop_rx_ba_session( sta, tid, WLAN_BACK_RECIPIENT, WLAN_REASON_QSTA_TIMEOUT); + } + mutex_unlock(&sta->ampdu_mlme.mtx); + spin_lock_bh(&sta->lock); + for (tid = 0; tid < STA_TID_NUM; tid++) { tid_tx = sta->ampdu_mlme.tid_tx[tid]; if (!tid_tx) continue; diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 9f00d3f..490be2f 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -740,7 +740,7 @@ static void ieee80211_iface_work(struct work_struct *work) mgmt->u.action.category == WLAN_CATEGORY_BACK) { int len = skb->len; - rcu_read_lock(); + mutex_lock(&local->sta_mtx); sta = sta_info_get(sdata, mgmt->sa); if (sta) { switch (mgmt->u.action.u.addba_req.action_code) { @@ -761,7 +761,7 @@ static void ieee80211_iface_work(struct work_struct *work) break; } } - rcu_read_unlock(); + mutex_unlock(&local->sta_mtx); } else if (ieee80211_is_data_qos(mgmt->frame_control)) { struct ieee80211_hdr *hdr = (void *)mgmt; /* @@ -781,7 +781,7 @@ static void ieee80211_iface_work(struct work_struct *work) * a block-ack session was active. That cannot be * right, so terminate the session. */ - rcu_read_lock(); + mutex_lock(&local->sta_mtx); sta = sta_info_get(sdata, mgmt->sa); if (sta) { u16 tid = *ieee80211_get_qos_ctl(hdr) & @@ -791,7 +791,7 @@ static void ieee80211_iface_work(struct work_struct *work) sta, tid, WLAN_BACK_RECIPIENT, WLAN_REASON_QSTA_REQUIRE_SETUP); } - rcu_read_unlock(); + mutex_unlock(&local->sta_mtx); } else switch (sdata->vif.type) { case NL80211_IFTYPE_STATION: ieee80211_sta_rx_queued_mgmt(sdata, skb); diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c index bd726c6..f54d8ba 100644 --- a/net/mac80211/sta_info.c +++ b/net/mac80211/sta_info.c @@ -236,6 +236,7 @@ struct sta_info *sta_info_alloc(struct ieee80211_sub_if_data *sdata, spin_lock_init(&sta->flaglock); INIT_WORK(&sta->drv_unblock_wk, sta_unblock); INIT_WORK(&sta->ampdu_mlme.work, ieee80211_ba_session_work); + mutex_init(&sta->ampdu_mlme.mtx); memcpy(sta->sta.addr, addr, ETH_ALEN); sta->local = local; diff --git a/net/mac80211/sta_info.h b/net/mac80211/sta_info.h index 500bafe..10d0fcb 100644 --- a/net/mac80211/sta_info.h +++ b/net/mac80211/sta_info.h @@ -142,8 +142,11 @@ struct tid_ampdu_rx { * @work: work struct for starting/stopping aggregation * @tid_rx_timer_expired: bitmap indicating on which TIDs the * RX timer expired until the work for it runs + * @mtx: mutex to protect all TX data (except non-NULL assignments + * to tid_tx[idx], which are protected by the sta spinlock) */ struct sta_ampdu_mlme { + struct mutex mtx; /* rx */ struct tid_ampdu_rx *tid_rx[STA_TID_NUM]; unsigned long tid_rx_timer_expired[BITS_TO_LONGS(STA_TID_NUM)]; -- cgit v1.1 From 83a5cbf73a13d0c8be019b22afec4407e4285aed Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:47 +0200 Subject: mac80211: defer TX agg session teardown to work Since we want the code to be able to sleep in the future, it must not be called from the timer directly. To achieve that, simply call the function drivers would call, and also use RCU in the timer to get the struct so we don't need to rely on the spinlock in the future. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index 9a00a79..0026604 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -200,11 +200,11 @@ static void sta_addba_resp_timer_expired(unsigned long data) struct tid_ampdu_tx *tid_tx; /* check if the TID waits for addBA response */ - spin_lock_bh(&sta->lock); - tid_tx = sta->ampdu_mlme.tid_tx[tid]; + rcu_read_lock(); + tid_tx = rcu_dereference(sta->ampdu_mlme.tid_tx[tid]); if (!tid_tx || test_bit(HT_AGG_STATE_RESPONSE_RECEIVED, &tid_tx->state)) { - spin_unlock_bh(&sta->lock); + rcu_read_unlock(); #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "timer expired on tid %d but we are not " "(or no longer) expecting addBA response there\n", @@ -217,8 +217,8 @@ static void sta_addba_resp_timer_expired(unsigned long data) printk(KERN_DEBUG "addBA response timer expired on tid %d\n", tid); #endif - ___ieee80211_stop_tx_ba_session(sta, tid, WLAN_BACK_INITIATOR); - spin_unlock_bh(&sta->lock); + ieee80211_stop_tx_ba_session(&sta->sta, tid); + rcu_read_unlock(); } static inline int ieee80211_ac_from_tid(int tid) -- cgit v1.1 From cfcdbde35e2b621cf56bedc38a3a81e8c28addb9 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:48 +0200 Subject: mac80211: change TX aggregation locking To prepare for allowing drivers to sleep in ampdu_action, change the locking in the TX aggregation code to use the mutex the RX part already uses. The spinlock is still necessary around some code to avoid races with TX, but now we can also synchronize_net() to avoid getting an inconsistent sequence number. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-tx.c | 94 ++++++++++++++++++++++++++--------------------- net/mac80211/driver-ops.h | 3 ++ net/mac80211/ht.c | 8 +--- 3 files changed, 58 insertions(+), 47 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index 0026604..5dff73e 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -6,7 +6,7 @@ * Copyright 2005-2006, Devicescape Software, Inc. * Copyright 2006-2007 Jiri Benc * Copyright 2007, Michael Wu - * Copyright 2007-2009, Intel Corporation + * Copyright 2007-2010, Intel Corporation * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -140,18 +140,23 @@ int ___ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, struct tid_ampdu_tx *tid_tx = sta->ampdu_mlme.tid_tx[tid]; int ret; - lockdep_assert_held(&sta->lock); + lockdep_assert_held(&sta->ampdu_mlme.mtx); - if (WARN_ON(!tid_tx)) + if (!tid_tx) return -ENOENT; + spin_lock_bh(&sta->lock); + if (test_bit(HT_AGG_STATE_WANT_START, &tid_tx->state)) { /* not even started yet! */ rcu_assign_pointer(sta->ampdu_mlme.tid_tx[tid], NULL); + spin_unlock_bh(&sta->lock); call_rcu(&tid_tx->rcu_head, kfree_tid_tx); return 0; } + spin_unlock_bh(&sta->lock); + #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "Tx BA session stop requested for %pM tid %u\n", sta->sta.addr, tid); @@ -269,6 +274,8 @@ void ieee80211_tx_ba_session_handle_start(struct sta_info *sta, int tid) u16 start_seq_num; int ret; + lockdep_assert_held(&sta->ampdu_mlme.mtx); + /* * While we're asking the driver about the aggregation, * stop the AC queue so that we don't have to worry @@ -281,10 +288,11 @@ void ieee80211_tx_ba_session_handle_start(struct sta_info *sta, int tid) clear_bit(HT_AGG_STATE_WANT_START, &tid_tx->state); /* - * This might be off by one due to a race that we can't - * really prevent here without synchronize_net() which - * can't be called now. + * make sure no packets are being processed to get + * valid starting sequence number */ + synchronize_net(); + start_seq_num = sta->tid_seq[tid] >> 4; ret = drv_ampdu_action(local, sdata, IEEE80211_AMPDU_TX_START, @@ -294,7 +302,10 @@ void ieee80211_tx_ba_session_handle_start(struct sta_info *sta, int tid) printk(KERN_DEBUG "BA request denied - HW unavailable for" " tid %d\n", tid); #endif + spin_lock_bh(&sta->lock); rcu_assign_pointer(sta->ampdu_mlme.tid_tx[tid], NULL); + spin_unlock_bh(&sta->lock); + ieee80211_wake_queue_agg(local, tid); call_rcu(&tid_tx->rcu_head, kfree_tid_tx); return; @@ -309,7 +320,9 @@ void ieee80211_tx_ba_session_handle_start(struct sta_info *sta, int tid) printk(KERN_DEBUG "activated addBA response timer on tid %d\n", tid); #endif + spin_lock_bh(&sta->lock); sta->ampdu_mlme.addba_req_num[tid]++; + spin_unlock_bh(&sta->lock); /* send AddBA request */ ieee80211_send_addba_request(sdata, sta->sta.addr, tid, @@ -445,16 +458,25 @@ ieee80211_agg_splice_finish(struct ieee80211_local *local, u16 tid) ieee80211_wake_queue_agg(local, tid); } -/* caller must hold sta->lock */ static void ieee80211_agg_tx_operational(struct ieee80211_local *local, struct sta_info *sta, u16 tid) { - lockdep_assert_held(&sta->lock); + lockdep_assert_held(&sta->ampdu_mlme.mtx); #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "Aggregation is on for tid %d\n", tid); #endif + drv_ampdu_action(local, sta->sdata, + IEEE80211_AMPDU_TX_OPERATIONAL, + &sta->sta, tid, NULL); + + /* + * synchronize with TX path, while splicing the TX path + * should block so it won't put more packets onto pending. + */ + spin_lock_bh(&sta->lock); + ieee80211_agg_splice_packets(local, sta->ampdu_mlme.tid_tx[tid], tid); /* * Now mark as operational. This will be visible @@ -464,9 +486,7 @@ static void ieee80211_agg_tx_operational(struct ieee80211_local *local, set_bit(HT_AGG_STATE_OPERATIONAL, &sta->ampdu_mlme.tid_tx[tid]->state); ieee80211_agg_splice_finish(local, tid); - drv_ampdu_action(local, sta->sdata, - IEEE80211_AMPDU_TX_OPERATIONAL, - &sta->sta, tid, NULL); + spin_unlock_bh(&sta->lock); } void ieee80211_start_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u16 tid) @@ -486,37 +506,35 @@ void ieee80211_start_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u16 tid) return; } - rcu_read_lock(); + mutex_lock(&local->sta_mtx); sta = sta_info_get(sdata, ra); if (!sta) { - rcu_read_unlock(); + mutex_unlock(&local->sta_mtx); #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "Could not find station: %pM\n", ra); #endif return; } - spin_lock_bh(&sta->lock); + mutex_lock(&sta->ampdu_mlme.mtx); tid_tx = sta->ampdu_mlme.tid_tx[tid]; if (WARN_ON(!tid_tx)) { #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "addBA was not requested!\n"); #endif - spin_unlock_bh(&sta->lock); - rcu_read_unlock(); - return; + goto unlock; } if (WARN_ON(test_and_set_bit(HT_AGG_STATE_DRV_READY, &tid_tx->state))) - goto out; + goto unlock; if (test_bit(HT_AGG_STATE_RESPONSE_RECEIVED, &tid_tx->state)) ieee80211_agg_tx_operational(local, sta, tid); - out: - spin_unlock_bh(&sta->lock); - rcu_read_unlock(); + unlock: + mutex_unlock(&sta->ampdu_mlme.mtx); + mutex_unlock(&local->sta_mtx); } void ieee80211_start_tx_ba_cb_irqsafe(struct ieee80211_vif *vif, @@ -548,21 +566,14 @@ EXPORT_SYMBOL(ieee80211_start_tx_ba_cb_irqsafe); int __ieee80211_stop_tx_ba_session(struct sta_info *sta, u16 tid, enum ieee80211_back_parties initiator) { - struct tid_ampdu_tx *tid_tx; int ret; - spin_lock_bh(&sta->lock); - tid_tx = sta->ampdu_mlme.tid_tx[tid]; - - if (!tid_tx) { - ret = -ENOENT; - goto unlock; - } + mutex_lock(&sta->ampdu_mlme.mtx); ret = ___ieee80211_stop_tx_ba_session(sta, tid, initiator); - unlock: - spin_unlock_bh(&sta->lock); + mutex_unlock(&sta->ampdu_mlme.mtx); + return ret; } @@ -627,16 +638,17 @@ void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid) ra, tid); #endif /* CONFIG_MAC80211_HT_DEBUG */ - rcu_read_lock(); + mutex_lock(&local->sta_mtx); + sta = sta_info_get(sdata, ra); if (!sta) { #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "Could not find station: %pM\n", ra); #endif - rcu_read_unlock(); - return; + goto unlock; } + mutex_lock(&sta->ampdu_mlme.mtx); spin_lock_bh(&sta->lock); tid_tx = sta->ampdu_mlme.tid_tx[tid]; @@ -644,9 +656,7 @@ void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid) #ifdef CONFIG_MAC80211_HT_DEBUG printk(KERN_DEBUG "unexpected callback to A-MPDU stop\n"); #endif - spin_unlock_bh(&sta->lock); - rcu_read_unlock(); - return; + goto unlock_sta; } if (tid_tx->stop_initiator == WLAN_BACK_INITIATOR) @@ -672,8 +682,11 @@ void ieee80211_stop_tx_ba_cb(struct ieee80211_vif *vif, u8 *ra, u8 tid) call_rcu(&tid_tx->rcu_head, kfree_tid_tx); + unlock_sta: spin_unlock_bh(&sta->lock); - rcu_read_unlock(); + mutex_unlock(&sta->ampdu_mlme.mtx); + unlock: + mutex_unlock(&local->sta_mtx); } void ieee80211_stop_tx_ba_cb_irqsafe(struct ieee80211_vif *vif, @@ -714,10 +727,9 @@ void ieee80211_process_addba_resp(struct ieee80211_local *local, capab = le16_to_cpu(mgmt->u.action.u.addba_resp.capab); tid = (capab & IEEE80211_ADDBA_PARAM_TID_MASK) >> 2; - spin_lock_bh(&sta->lock); + mutex_lock(&sta->ampdu_mlme.mtx); tid_tx = sta->ampdu_mlme.tid_tx[tid]; - if (!tid_tx) goto out; @@ -751,5 +763,5 @@ void ieee80211_process_addba_resp(struct ieee80211_local *local, } out: - spin_unlock_bh(&sta->lock); + mutex_unlock(&sta->ampdu_mlme.mtx); } diff --git a/net/mac80211/driver-ops.h b/net/mac80211/driver-ops.h index e5e7ef1..7e86c6f 100644 --- a/net/mac80211/driver-ops.h +++ b/net/mac80211/driver-ops.h @@ -349,6 +349,9 @@ static inline int drv_ampdu_action(struct ieee80211_local *local, u16 *ssn) { int ret = -EOPNOTSUPP; + + might_sleep(); + local_bh_disable(); if (local->ops->ampdu_action) ret = local->ops->ampdu_action(&local->hw, &sdata->vif, action, diff --git a/net/mac80211/ht.c b/net/mac80211/ht.c index e29be64..be928ef 100644 --- a/net/mac80211/ht.c +++ b/net/mac80211/ht.c @@ -6,7 +6,7 @@ * Copyright 2005-2006, Devicescape Software, Inc. * Copyright 2006-2007 Jiri Benc * Copyright 2007, Michael Wu - * Copyright 2007-2008, Intel Corporation + * Copyright 2007-2010, Intel Corporation * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -136,11 +136,7 @@ void ieee80211_ba_session_work(struct work_struct *work) ___ieee80211_stop_rx_ba_session( sta, tid, WLAN_BACK_RECIPIENT, WLAN_REASON_QSTA_TIMEOUT); - } - mutex_unlock(&sta->ampdu_mlme.mtx); - spin_lock_bh(&sta->lock); - for (tid = 0; tid < STA_TID_NUM; tid++) { tid_tx = sta->ampdu_mlme.tid_tx[tid]; if (!tid_tx) continue; @@ -152,7 +148,7 @@ void ieee80211_ba_session_work(struct work_struct *work) ___ieee80211_stop_tx_ba_session(sta, tid, WLAN_BACK_INITIATOR); } - spin_unlock_bh(&sta->lock); + mutex_unlock(&sta->ampdu_mlme.mtx); } void ieee80211_send_delba(struct ieee80211_sub_if_data *sdata, -- cgit v1.1 From 85ad181ea78861f69b007599cec9e6ba33fcdf8a Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:49 +0200 Subject: mac80211: allow drivers to sleep in ampdu_action Allow drivers to sleep, and indicate this in the documentation. ath9k has some locking I don't understand, so keep it safe and disable BHs in it, all other drivers look fine with the context change. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/driver-ops.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'net') diff --git a/net/mac80211/driver-ops.h b/net/mac80211/driver-ops.h index 7e86c6f..a4fcbcc 100644 --- a/net/mac80211/driver-ops.h +++ b/net/mac80211/driver-ops.h @@ -352,11 +352,10 @@ static inline int drv_ampdu_action(struct ieee80211_local *local, might_sleep(); - local_bh_disable(); if (local->ops->ampdu_action) ret = local->ops->ampdu_action(&local->hw, &sdata->vif, action, sta, tid, ssn); - local_bh_enable(); + trace_drv_ampdu_action(local, sdata, action, sta, tid, ssn, ret); return ret; } -- cgit v1.1 From 73a72a81d27b71f9ace31668d2dd7f3ac1c8228e Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:50 +0200 Subject: mac80211: update aggregation documentation Even before the recent changes, the documentation for TX aggregation was somewhat out of date. Update it and also add documentation for the RX side. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/agg-rx.c | 23 +++++++++++++++++++++++ net/mac80211/agg-tx.c | 43 +++++++++++++++++++++++++++---------------- 2 files changed, 50 insertions(+), 16 deletions(-) (limited to 'net') diff --git a/net/mac80211/agg-rx.c b/net/mac80211/agg-rx.c index a843df2..965b272 100644 --- a/net/mac80211/agg-rx.c +++ b/net/mac80211/agg-rx.c @@ -13,6 +13,29 @@ * published by the Free Software Foundation. */ +/** + * DOC: RX A-MPDU aggregation + * + * Aggregation on the RX side requires only implementing the + * @ampdu_action callback that is invoked to start/stop any + * block-ack sessions for RX aggregation. + * + * When RX aggregation is started by the peer, the driver is + * notified via @ampdu_action function, with the + * %IEEE80211_AMPDU_RX_START action, and may reject the request + * in which case a negative response is sent to the peer, if it + * accepts it a positive response is sent. + * + * While the session is active, the device/driver are required + * to de-aggregate frames and pass them up one by one to mac80211, + * which will handle the reorder buffer. + * + * When the aggregation session is stopped again by the peer or + * ourselves, the driver's @ampdu_action function will be called + * with the action %IEEE80211_AMPDU_RX_STOP. In this case, the + * call must not fail. + */ + #include #include #include diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index 5dff73e..c893f23 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -21,28 +21,39 @@ #include "wme.h" /** - * DOC: TX aggregation + * DOC: TX A-MPDU aggregation * * Aggregation on the TX side requires setting the hardware flag - * %IEEE80211_HW_AMPDU_AGGREGATION as well as, if present, the @ampdu_queues - * hardware parameter to the number of hardware AMPDU queues. If there are no - * hardware queues then the driver will (currently) have to do all frame - * buffering. + * %IEEE80211_HW_AMPDU_AGGREGATION. The driver will then be handed + * packets with a flag indicating A-MPDU aggregation. The driver + * or device is responsible for actually aggregating the frames, + * as well as deciding how many and which to aggregate. * - * When TX aggregation is started by some subsystem (usually the rate control - * algorithm would be appropriate) by calling the - * ieee80211_start_tx_ba_session() function, the driver will be notified via - * its @ampdu_action function, with the %IEEE80211_AMPDU_TX_START action. + * When TX aggregation is started by some subsystem (usually the rate + * control algorithm would be appropriate) by calling the + * ieee80211_start_tx_ba_session() function, the driver will be + * notified via its @ampdu_action function, with the + * %IEEE80211_AMPDU_TX_START action. * * In response to that, the driver is later required to call the - * ieee80211_start_tx_ba_cb() (or ieee80211_start_tx_ba_cb_irqsafe()) - * function, which will start the aggregation session. + * ieee80211_start_tx_ba_cb_irqsafe() function, which will really + * start the aggregation session after the peer has also responded. + * If the peer responds negatively, the session will be stopped + * again right away. Note that it is possible for the aggregation + * session to be stopped before the driver has indicated that it + * is done setting it up, in which case it must not indicate the + * setup completion. * - * Similarly, when the aggregation session is stopped by - * ieee80211_stop_tx_ba_session(), the driver's @ampdu_action function will - * be called with the action %IEEE80211_AMPDU_TX_STOP. In this case, the - * call must not fail, and the driver must later call ieee80211_stop_tx_ba_cb() - * (or ieee80211_stop_tx_ba_cb_irqsafe()). + * Also note that, since we also need to wait for a response from + * the peer, the driver is notified of the completion of the + * handshake by the %IEEE80211_AMPDU_TX_OPERATIONAL action to the + * @ampdu_action callback. + * + * Similarly, when the aggregation session is stopped by the peer + * or something calling ieee80211_stop_tx_ba_session(), the driver's + * @ampdu_action function will be called with the action + * %IEEE80211_AMPDU_TX_STOP. In this case, the call must not fail, + * and the driver must later call ieee80211_stop_tx_ba_cb_irqsafe(). */ static void ieee80211_send_addba_request(struct ieee80211_sub_if_data *sdata, -- cgit v1.1 From 8b58ff832df9222e1fb3c58ae7f745f95569b55e Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:21:51 +0200 Subject: mac80211: fix mgmt frame accounting The recent change to processing action frames from the management frame queue had already broken action frame accounting, and my rework didn't help either. So add back accounting and simplify the code with a label rather than duplicating it, and also add accounting for management frames. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/rx.c | 25 +++++++++++++------------ 1 file changed, 13 insertions(+), 12 deletions(-) (limited to 'net') diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index ee01dac..1594ebe 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -1966,10 +1966,7 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) goto invalid; } - rx->skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; - skb_queue_tail(&sdata->skb_queue, rx->skb); - ieee80211_queue_work(&local->hw, &sdata->work); - return RX_QUEUED; + goto queue; case WLAN_CATEGORY_SPECTRUM_MGMT: if (local->hw.conf.channel->band != IEEE80211_BAND_5GHZ) break; @@ -1999,10 +1996,7 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) if (memcmp(mgmt->bssid, sdata->u.mgd.bssid, ETH_ALEN)) break; - rx->skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; - skb_queue_tail(&sdata->skb_queue, rx->skb); - ieee80211_queue_work(&local->hw, &sdata->work); - return RX_QUEUED; + goto queue; } break; case WLAN_CATEGORY_SA_QUERY: @@ -2022,10 +2016,7 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) case WLAN_CATEGORY_MESH_PATH_SEL: if (!ieee80211_vif_is_mesh(&sdata->vif)) break; - rx->skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; - skb_queue_tail(&sdata->skb_queue, rx->skb); - ieee80211_queue_work(&local->hw, &sdata->work); - return RX_QUEUED; + goto queue; } invalid: @@ -2076,6 +2067,14 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) rx->sta->rx_packets++; dev_kfree_skb(rx->skb); return RX_QUEUED; + + queue: + rx->skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; + skb_queue_tail(&sdata->skb_queue, rx->skb); + ieee80211_queue_work(&local->hw, &sdata->work); + if (rx->sta) + rx->sta->rx_packets++; + return RX_QUEUED; } static ieee80211_rx_result debug_noinline @@ -2131,6 +2130,8 @@ ieee80211_rx_h_mgmt(struct ieee80211_rx_data *rx) rx->skb->pkt_type = IEEE80211_SDATA_QUEUE_TYPE_FRAME; skb_queue_tail(&sdata->skb_queue, rx->skb); ieee80211_queue_work(&rx->local->hw, &sdata->work); + if (rx->sta) + rx->sta->rx_packets++; return RX_QUEUED; } -- cgit v1.1 From 4efc76bdbf9bfd2b8624c4dd52b9d8ffed176b31 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 10 Jun 2010 10:56:20 +0200 Subject: mac80211: bracket driver tracing Currently, driver tracing is sometimes invoked after and sometimes before the actual driver callback. This is fine as long as the driver has no tracing itself, but as soon as it does it gets confusing. To make traces containing such information easier to read, introduce a return tracer in mac80211 that essentially brackets any driver tracing, and invoke the real trace before the driver's callback, only showing the return value, if any, afterwards. Since tracing records the process, there's no problem with overlapping calls if that should happen. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/driver-ops.h | 89 ++++++++++++++------- net/mac80211/driver-trace.h | 189 +++++++++++++++++++++++--------------------- 2 files changed, 156 insertions(+), 122 deletions(-) (limited to 'net') diff --git a/net/mac80211/driver-ops.h b/net/mac80211/driver-ops.h index a4fcbcc..965d64f 100644 --- a/net/mac80211/driver-ops.h +++ b/net/mac80211/driver-ops.h @@ -16,10 +16,11 @@ static inline int drv_start(struct ieee80211_local *local) might_sleep(); + trace_drv_start(local); local->started = true; smp_mb(); ret = local->ops->start(&local->hw); - trace_drv_start(local, ret); + trace_drv_return_int(local, ret); return ret; } @@ -27,8 +28,9 @@ static inline void drv_stop(struct ieee80211_local *local) { might_sleep(); - local->ops->stop(&local->hw); trace_drv_stop(local); + local->ops->stop(&local->hw); + trace_drv_return_void(local); /* sync away all work on the tasklet before clearing started */ tasklet_disable(&local->tasklet); @@ -46,8 +48,9 @@ static inline int drv_add_interface(struct ieee80211_local *local, might_sleep(); + trace_drv_add_interface(local, vif_to_sdata(vif)); ret = local->ops->add_interface(&local->hw, vif); - trace_drv_add_interface(local, vif_to_sdata(vif), ret); + trace_drv_return_int(local, ret); return ret; } @@ -56,8 +59,9 @@ static inline void drv_remove_interface(struct ieee80211_local *local, { might_sleep(); - local->ops->remove_interface(&local->hw, vif); trace_drv_remove_interface(local, vif_to_sdata(vif)); + local->ops->remove_interface(&local->hw, vif); + trace_drv_return_void(local); } static inline int drv_config(struct ieee80211_local *local, u32 changed) @@ -66,8 +70,9 @@ static inline int drv_config(struct ieee80211_local *local, u32 changed) might_sleep(); + trace_drv_config(local, changed); ret = local->ops->config(&local->hw, changed); - trace_drv_config(local, changed, ret); + trace_drv_return_int(local, ret); return ret; } @@ -78,9 +83,10 @@ static inline void drv_bss_info_changed(struct ieee80211_local *local, { might_sleep(); + trace_drv_bss_info_changed(local, sdata, info, changed); if (local->ops->bss_info_changed) local->ops->bss_info_changed(&local->hw, &sdata->vif, info, changed); - trace_drv_bss_info_changed(local, sdata, info, changed); + trace_drv_return_void(local); } struct in_ifaddr; @@ -92,11 +98,11 @@ static inline int drv_configure_arp_filter(struct ieee80211_local *local, might_sleep(); + trace_drv_configure_arp_filter(local, vif_to_sdata(vif)); if (local->ops->configure_arp_filter) ret = local->ops->configure_arp_filter(&local->hw, vif, ifa_list); - - trace_drv_configure_arp_filter(local, vif_to_sdata(vif), ifa_list, ret); + trace_drv_return_int(local, ret); return ret; } @@ -105,10 +111,12 @@ static inline u64 drv_prepare_multicast(struct ieee80211_local *local, { u64 ret = 0; + trace_drv_prepare_multicast(local, mc_list->count); + if (local->ops->prepare_multicast) ret = local->ops->prepare_multicast(&local->hw, mc_list); - trace_drv_prepare_multicast(local, mc_list->count, ret); + trace_drv_return_u64(local, ret); return ret; } @@ -120,19 +128,21 @@ static inline void drv_configure_filter(struct ieee80211_local *local, { might_sleep(); - local->ops->configure_filter(&local->hw, changed_flags, total_flags, - multicast); trace_drv_configure_filter(local, changed_flags, total_flags, multicast); + local->ops->configure_filter(&local->hw, changed_flags, total_flags, + multicast); + trace_drv_return_void(local); } static inline int drv_set_tim(struct ieee80211_local *local, struct ieee80211_sta *sta, bool set) { int ret = 0; + trace_drv_set_tim(local, sta, set); if (local->ops->set_tim) ret = local->ops->set_tim(&local->hw, sta, set); - trace_drv_set_tim(local, sta, set, ret); + trace_drv_return_int(local, ret); return ret; } @@ -146,8 +156,9 @@ static inline int drv_set_key(struct ieee80211_local *local, might_sleep(); + trace_drv_set_key(local, cmd, sdata, sta, key); ret = local->ops->set_key(&local->hw, cmd, &sdata->vif, sta, key); - trace_drv_set_key(local, cmd, sdata, sta, key, ret); + trace_drv_return_int(local, ret); return ret; } @@ -162,10 +173,11 @@ static inline void drv_update_tkip_key(struct ieee80211_local *local, if (sta) ista = &sta->sta; + trace_drv_update_tkip_key(local, sdata, conf, ista, iv32); if (local->ops->update_tkip_key) local->ops->update_tkip_key(&local->hw, &sdata->vif, conf, ista, iv32, phase1key); - trace_drv_update_tkip_key(local, sdata, conf, ista, iv32); + trace_drv_return_void(local); } static inline int drv_hw_scan(struct ieee80211_local *local, @@ -176,8 +188,9 @@ static inline int drv_hw_scan(struct ieee80211_local *local, might_sleep(); + trace_drv_hw_scan(local, sdata, req); ret = local->ops->hw_scan(&local->hw, &sdata->vif, req); - trace_drv_hw_scan(local, sdata, req, ret); + trace_drv_return_int(local, ret); return ret; } @@ -185,18 +198,20 @@ static inline void drv_sw_scan_start(struct ieee80211_local *local) { might_sleep(); + trace_drv_sw_scan_start(local); if (local->ops->sw_scan_start) local->ops->sw_scan_start(&local->hw); - trace_drv_sw_scan_start(local); + trace_drv_return_void(local); } static inline void drv_sw_scan_complete(struct ieee80211_local *local) { might_sleep(); + trace_drv_sw_scan_complete(local); if (local->ops->sw_scan_complete) local->ops->sw_scan_complete(&local->hw); - trace_drv_sw_scan_complete(local); + trace_drv_return_void(local); } static inline int drv_get_stats(struct ieee80211_local *local, @@ -228,9 +243,10 @@ static inline int drv_set_rts_threshold(struct ieee80211_local *local, might_sleep(); + trace_drv_set_rts_threshold(local, value); if (local->ops->set_rts_threshold) ret = local->ops->set_rts_threshold(&local->hw, value); - trace_drv_set_rts_threshold(local, value, ret); + trace_drv_return_int(local, ret); return ret; } @@ -240,12 +256,13 @@ static inline int drv_set_coverage_class(struct ieee80211_local *local, int ret = 0; might_sleep(); + trace_drv_set_coverage_class(local, value); if (local->ops->set_coverage_class) local->ops->set_coverage_class(&local->hw, value); else ret = -EOPNOTSUPP; - trace_drv_set_coverage_class(local, value, ret); + trace_drv_return_int(local, ret); return ret; } @@ -254,9 +271,10 @@ static inline void drv_sta_notify(struct ieee80211_local *local, enum sta_notify_cmd cmd, struct ieee80211_sta *sta) { + trace_drv_sta_notify(local, sdata, cmd, sta); if (local->ops->sta_notify) local->ops->sta_notify(&local->hw, &sdata->vif, cmd, sta); - trace_drv_sta_notify(local, sdata, cmd, sta); + trace_drv_return_void(local); } static inline int drv_sta_add(struct ieee80211_local *local, @@ -267,10 +285,11 @@ static inline int drv_sta_add(struct ieee80211_local *local, might_sleep(); + trace_drv_sta_add(local, sdata, sta); if (local->ops->sta_add) ret = local->ops->sta_add(&local->hw, &sdata->vif, sta); - trace_drv_sta_add(local, sdata, sta, ret); + trace_drv_return_int(local, ret); return ret; } @@ -281,10 +300,11 @@ static inline void drv_sta_remove(struct ieee80211_local *local, { might_sleep(); + trace_drv_sta_remove(local, sdata, sta); if (local->ops->sta_remove) local->ops->sta_remove(&local->hw, &sdata->vif, sta); - trace_drv_sta_remove(local, sdata, sta); + trace_drv_return_void(local); } static inline int drv_conf_tx(struct ieee80211_local *local, u16 queue, @@ -294,9 +314,10 @@ static inline int drv_conf_tx(struct ieee80211_local *local, u16 queue, might_sleep(); + trace_drv_conf_tx(local, queue, params); if (local->ops->conf_tx) ret = local->ops->conf_tx(&local->hw, queue, params); - trace_drv_conf_tx(local, queue, params, ret); + trace_drv_return_int(local, ret); return ret; } @@ -306,9 +327,10 @@ static inline u64 drv_get_tsf(struct ieee80211_local *local) might_sleep(); + trace_drv_get_tsf(local); if (local->ops->get_tsf) ret = local->ops->get_tsf(&local->hw); - trace_drv_get_tsf(local, ret); + trace_drv_return_u64(local, ret); return ret; } @@ -316,18 +338,20 @@ static inline void drv_set_tsf(struct ieee80211_local *local, u64 tsf) { might_sleep(); + trace_drv_set_tsf(local, tsf); if (local->ops->set_tsf) local->ops->set_tsf(&local->hw, tsf); - trace_drv_set_tsf(local, tsf); + trace_drv_return_void(local); } static inline void drv_reset_tsf(struct ieee80211_local *local) { might_sleep(); + trace_drv_reset_tsf(local); if (local->ops->reset_tsf) local->ops->reset_tsf(&local->hw); - trace_drv_reset_tsf(local); + trace_drv_return_void(local); } static inline int drv_tx_last_beacon(struct ieee80211_local *local) @@ -336,9 +360,10 @@ static inline int drv_tx_last_beacon(struct ieee80211_local *local) might_sleep(); + trace_drv_tx_last_beacon(local); if (local->ops->tx_last_beacon) ret = local->ops->tx_last_beacon(&local->hw); - trace_drv_tx_last_beacon(local, ret); + trace_drv_return_int(local, ret); return ret; } @@ -352,11 +377,14 @@ static inline int drv_ampdu_action(struct ieee80211_local *local, might_sleep(); + trace_drv_ampdu_action(local, sdata, action, sta, tid, ssn); + if (local->ops->ampdu_action) ret = local->ops->ampdu_action(&local->hw, &sdata->vif, action, sta, tid, ssn); - trace_drv_ampdu_action(local, sdata, action, sta, tid, ssn, ret); + trace_drv_return_int(local, ret); + return ret; } @@ -385,6 +413,7 @@ static inline void drv_flush(struct ieee80211_local *local, bool drop) trace_drv_flush(local, drop); if (local->ops->flush) local->ops->flush(&local->hw, drop); + trace_drv_return_void(local); } static inline void drv_channel_switch(struct ieee80211_local *local, @@ -392,9 +421,9 @@ static inline void drv_channel_switch(struct ieee80211_local *local, { might_sleep(); - local->ops->channel_switch(&local->hw, ch_switch); - trace_drv_channel_switch(local, ch_switch); + local->ops->channel_switch(&local->hw, ch_switch); + trace_drv_return_void(local); } #endif /* __MAC80211_DRIVER_OPS */ diff --git a/net/mac80211/driver-trace.h b/net/mac80211/driver-trace.h index 6b906301..06444ea 100644 --- a/net/mac80211/driver-trace.h +++ b/net/mac80211/driver-trace.h @@ -36,20 +36,58 @@ static inline void trace_ ## name(proto) {} * Tracing for driver callbacks. */ -TRACE_EVENT(drv_start, - TP_PROTO(struct ieee80211_local *local, int ret), +TRACE_EVENT(drv_return_void, + TP_PROTO(struct ieee80211_local *local), + TP_ARGS(local), + TP_STRUCT__entry( + LOCAL_ENTRY + ), + TP_fast_assign( + LOCAL_ASSIGN; + ), + TP_printk(LOCAL_PR_FMT, LOCAL_PR_ARG) +); +TRACE_EVENT(drv_return_int, + TP_PROTO(struct ieee80211_local *local, int ret), TP_ARGS(local, ret), - TP_STRUCT__entry( LOCAL_ENTRY __field(int, ret) ), + TP_fast_assign( + LOCAL_ASSIGN; + __entry->ret = ret; + ), + TP_printk(LOCAL_PR_FMT " - %d", LOCAL_PR_ARG, __entry->ret) +); +TRACE_EVENT(drv_return_u64, + TP_PROTO(struct ieee80211_local *local, u64 ret), + TP_ARGS(local, ret), + TP_STRUCT__entry( + LOCAL_ENTRY + __field(u64, ret) + ), TP_fast_assign( LOCAL_ASSIGN; __entry->ret = ret; ), + TP_printk(LOCAL_PR_FMT " - %llu", LOCAL_PR_ARG, __entry->ret) +); + +TRACE_EVENT(drv_start, + TP_PROTO(struct ieee80211_local *local), + + TP_ARGS(local), + + TP_STRUCT__entry( + LOCAL_ENTRY + ), + + TP_fast_assign( + LOCAL_ASSIGN; + ), TP_printk( LOCAL_PR_FMT, LOCAL_PR_ARG @@ -76,28 +114,25 @@ TRACE_EVENT(drv_stop, TRACE_EVENT(drv_add_interface, TP_PROTO(struct ieee80211_local *local, - struct ieee80211_sub_if_data *sdata, - int ret), + struct ieee80211_sub_if_data *sdata), - TP_ARGS(local, sdata, ret), + TP_ARGS(local, sdata), TP_STRUCT__entry( LOCAL_ENTRY VIF_ENTRY __array(char, addr, 6) - __field(int, ret) ), TP_fast_assign( LOCAL_ASSIGN; VIF_ASSIGN; memcpy(__entry->addr, sdata->vif.addr, 6); - __entry->ret = ret; ), TP_printk( - LOCAL_PR_FMT VIF_PR_FMT " addr:%pM ret:%d", - LOCAL_PR_ARG, VIF_PR_ARG, __entry->addr, __entry->ret + LOCAL_PR_FMT VIF_PR_FMT " addr:%pM", + LOCAL_PR_ARG, VIF_PR_ARG, __entry->addr ) ); @@ -126,15 +161,13 @@ TRACE_EVENT(drv_remove_interface, TRACE_EVENT(drv_config, TP_PROTO(struct ieee80211_local *local, - u32 changed, - int ret), + u32 changed), - TP_ARGS(local, changed, ret), + TP_ARGS(local, changed), TP_STRUCT__entry( LOCAL_ENTRY __field(u32, changed) - __field(int, ret) __field(u32, flags) __field(int, power_level) __field(int, dynamic_ps_timeout) @@ -150,7 +183,6 @@ TRACE_EVENT(drv_config, TP_fast_assign( LOCAL_ASSIGN; __entry->changed = changed; - __entry->ret = ret; __entry->flags = local->hw.conf.flags; __entry->power_level = local->hw.conf.power_level; __entry->dynamic_ps_timeout = local->hw.conf.dynamic_ps_timeout; @@ -164,8 +196,8 @@ TRACE_EVENT(drv_config, ), TP_printk( - LOCAL_PR_FMT " ch:%#x freq:%d ret:%d", - LOCAL_PR_ARG, __entry->changed, __entry->center_freq, __entry->ret + LOCAL_PR_FMT " ch:%#x freq:%d", + LOCAL_PR_ARG, __entry->changed, __entry->center_freq ) ); @@ -221,50 +253,44 @@ TRACE_EVENT(drv_bss_info_changed, TRACE_EVENT(drv_configure_arp_filter, TP_PROTO(struct ieee80211_local *local, - struct ieee80211_sub_if_data *sdata, - struct in_ifaddr *ifa_list, int ret), + struct ieee80211_sub_if_data *sdata), - TP_ARGS(local, sdata, ifa_list, ret), + TP_ARGS(local, sdata), TP_STRUCT__entry( LOCAL_ENTRY VIF_ENTRY - __field(int, ret) ), TP_fast_assign( LOCAL_ASSIGN; VIF_ASSIGN; - __entry->ret = ret; ), TP_printk( - VIF_PR_FMT LOCAL_PR_FMT " ret:%d", - VIF_PR_ARG, LOCAL_PR_ARG, __entry->ret + VIF_PR_FMT LOCAL_PR_FMT, + VIF_PR_ARG, LOCAL_PR_ARG ) ); TRACE_EVENT(drv_prepare_multicast, - TP_PROTO(struct ieee80211_local *local, int mc_count, u64 ret), + TP_PROTO(struct ieee80211_local *local, int mc_count), - TP_ARGS(local, mc_count, ret), + TP_ARGS(local, mc_count), TP_STRUCT__entry( LOCAL_ENTRY __field(int, mc_count) - __field(u64, ret) ), TP_fast_assign( LOCAL_ASSIGN; __entry->mc_count = mc_count; - __entry->ret = ret; ), TP_printk( - LOCAL_PR_FMT " prepare mc (%d): %llx", - LOCAL_PR_ARG, __entry->mc_count, - (unsigned long long) __entry->ret + LOCAL_PR_FMT " prepare mc (%d)", + LOCAL_PR_ARG, __entry->mc_count ) ); @@ -298,27 +324,25 @@ TRACE_EVENT(drv_configure_filter, TRACE_EVENT(drv_set_tim, TP_PROTO(struct ieee80211_local *local, - struct ieee80211_sta *sta, bool set, int ret), + struct ieee80211_sta *sta, bool set), - TP_ARGS(local, sta, set, ret), + TP_ARGS(local, sta, set), TP_STRUCT__entry( LOCAL_ENTRY STA_ENTRY __field(bool, set) - __field(int, ret) ), TP_fast_assign( LOCAL_ASSIGN; STA_ASSIGN; __entry->set = set; - __entry->ret = ret; ), TP_printk( - LOCAL_PR_FMT STA_PR_FMT " set:%d ret:%d", - LOCAL_PR_ARG, STA_PR_FMT, __entry->set, __entry->ret + LOCAL_PR_FMT STA_PR_FMT " set:%d", + LOCAL_PR_ARG, STA_PR_FMT, __entry->set ) ); @@ -326,9 +350,9 @@ TRACE_EVENT(drv_set_key, TP_PROTO(struct ieee80211_local *local, enum set_key_cmd cmd, struct ieee80211_sub_if_data *sdata, struct ieee80211_sta *sta, - struct ieee80211_key_conf *key, int ret), + struct ieee80211_key_conf *key), - TP_ARGS(local, cmd, sdata, sta, key, ret), + TP_ARGS(local, cmd, sdata, sta, key), TP_STRUCT__entry( LOCAL_ENTRY @@ -338,7 +362,6 @@ TRACE_EVENT(drv_set_key, __field(u8, hw_key_idx) __field(u8, flags) __field(s8, keyidx) - __field(int, ret) ), TP_fast_assign( @@ -349,12 +372,11 @@ TRACE_EVENT(drv_set_key, __entry->flags = key->flags; __entry->keyidx = key->keyidx; __entry->hw_key_idx = key->hw_key_idx; - __entry->ret = ret; ), TP_printk( - LOCAL_PR_FMT VIF_PR_FMT STA_PR_FMT " ret:%d", - LOCAL_PR_ARG, VIF_PR_ARG, STA_PR_ARG, __entry->ret + LOCAL_PR_FMT VIF_PR_FMT STA_PR_FMT, + LOCAL_PR_ARG, VIF_PR_ARG, STA_PR_ARG ) ); @@ -389,25 +411,23 @@ TRACE_EVENT(drv_update_tkip_key, TRACE_EVENT(drv_hw_scan, TP_PROTO(struct ieee80211_local *local, struct ieee80211_sub_if_data *sdata, - struct cfg80211_scan_request *req, int ret), + struct cfg80211_scan_request *req), - TP_ARGS(local, sdata, req, ret), + TP_ARGS(local, sdata, req), TP_STRUCT__entry( LOCAL_ENTRY VIF_ENTRY - __field(int, ret) ), TP_fast_assign( LOCAL_ASSIGN; VIF_ASSIGN; - __entry->ret = ret; ), TP_printk( - LOCAL_PR_FMT VIF_PR_FMT " ret:%d", - LOCAL_PR_ARG,VIF_PR_ARG, __entry->ret + LOCAL_PR_FMT VIF_PR_FMT, + LOCAL_PR_ARG,VIF_PR_ARG ) ); @@ -504,48 +524,44 @@ TRACE_EVENT(drv_get_tkip_seq, ); TRACE_EVENT(drv_set_rts_threshold, - TP_PROTO(struct ieee80211_local *local, u32 value, int ret), + TP_PROTO(struct ieee80211_local *local, u32 value), - TP_ARGS(local, value, ret), + TP_ARGS(local, value), TP_STRUCT__entry( LOCAL_ENTRY __field(u32, value) - __field(int, ret) ), TP_fast_assign( LOCAL_ASSIGN; - __entry->ret = ret; __entry->value = value; ), TP_printk( - LOCAL_PR_FMT " value:%d ret:%d", - LOCAL_PR_ARG, __entry->value, __entry->ret + LOCAL_PR_FMT " value:%d", + LOCAL_PR_ARG, __entry->value ) ); TRACE_EVENT(drv_set_coverage_class, - TP_PROTO(struct ieee80211_local *local, u8 value, int ret), + TP_PROTO(struct ieee80211_local *local, u8 value), - TP_ARGS(local, value, ret), + TP_ARGS(local, value), TP_STRUCT__entry( LOCAL_ENTRY __field(u8, value) - __field(int, ret) ), TP_fast_assign( LOCAL_ASSIGN; - __entry->ret = ret; __entry->value = value; ), TP_printk( - LOCAL_PR_FMT " value:%d ret:%d", - LOCAL_PR_ARG, __entry->value, __entry->ret + LOCAL_PR_FMT " value:%d", + LOCAL_PR_ARG, __entry->value ) ); @@ -580,27 +596,25 @@ TRACE_EVENT(drv_sta_notify, TRACE_EVENT(drv_sta_add, TP_PROTO(struct ieee80211_local *local, struct ieee80211_sub_if_data *sdata, - struct ieee80211_sta *sta, int ret), + struct ieee80211_sta *sta), - TP_ARGS(local, sdata, sta, ret), + TP_ARGS(local, sdata, sta), TP_STRUCT__entry( LOCAL_ENTRY VIF_ENTRY STA_ENTRY - __field(int, ret) ), TP_fast_assign( LOCAL_ASSIGN; VIF_ASSIGN; STA_ASSIGN; - __entry->ret = ret; ), TP_printk( - LOCAL_PR_FMT VIF_PR_FMT STA_PR_FMT " ret:%d", - LOCAL_PR_ARG, VIF_PR_ARG, STA_PR_ARG, __entry->ret + LOCAL_PR_FMT VIF_PR_FMT STA_PR_FMT, + LOCAL_PR_ARG, VIF_PR_ARG, STA_PR_ARG ) ); @@ -631,10 +645,9 @@ TRACE_EVENT(drv_sta_remove, TRACE_EVENT(drv_conf_tx, TP_PROTO(struct ieee80211_local *local, u16 queue, - const struct ieee80211_tx_queue_params *params, - int ret), + const struct ieee80211_tx_queue_params *params), - TP_ARGS(local, queue, params, ret), + TP_ARGS(local, queue, params), TP_STRUCT__entry( LOCAL_ENTRY @@ -643,13 +656,11 @@ TRACE_EVENT(drv_conf_tx, __field(u16, cw_min) __field(u16, cw_max) __field(u8, aifs) - __field(int, ret) ), TP_fast_assign( LOCAL_ASSIGN; __entry->queue = queue; - __entry->ret = ret; __entry->txop = params->txop; __entry->cw_max = params->cw_max; __entry->cw_min = params->cw_min; @@ -657,29 +668,27 @@ TRACE_EVENT(drv_conf_tx, ), TP_printk( - LOCAL_PR_FMT " queue:%d ret:%d", - LOCAL_PR_ARG, __entry->queue, __entry->ret + LOCAL_PR_FMT " queue:%d", + LOCAL_PR_ARG, __entry->queue ) ); TRACE_EVENT(drv_get_tsf, - TP_PROTO(struct ieee80211_local *local, u64 ret), + TP_PROTO(struct ieee80211_local *local), - TP_ARGS(local, ret), + TP_ARGS(local), TP_STRUCT__entry( LOCAL_ENTRY - __field(u64, ret) ), TP_fast_assign( LOCAL_ASSIGN; - __entry->ret = ret; ), TP_printk( - LOCAL_PR_FMT " ret:%llu", - LOCAL_PR_ARG, (unsigned long long)__entry->ret + LOCAL_PR_FMT, + LOCAL_PR_ARG ) ); @@ -723,23 +732,21 @@ TRACE_EVENT(drv_reset_tsf, ); TRACE_EVENT(drv_tx_last_beacon, - TP_PROTO(struct ieee80211_local *local, int ret), + TP_PROTO(struct ieee80211_local *local), - TP_ARGS(local, ret), + TP_ARGS(local), TP_STRUCT__entry( LOCAL_ENTRY - __field(int, ret) ), TP_fast_assign( LOCAL_ASSIGN; - __entry->ret = ret; ), TP_printk( - LOCAL_PR_FMT " ret:%d", - LOCAL_PR_ARG, __entry->ret + LOCAL_PR_FMT, + LOCAL_PR_ARG ) ); @@ -748,9 +755,9 @@ TRACE_EVENT(drv_ampdu_action, struct ieee80211_sub_if_data *sdata, enum ieee80211_ampdu_mlme_action action, struct ieee80211_sta *sta, u16 tid, - u16 *ssn, int ret), + u16 *ssn), - TP_ARGS(local, sdata, action, sta, tid, ssn, ret), + TP_ARGS(local, sdata, action, sta, tid, ssn), TP_STRUCT__entry( LOCAL_ENTRY @@ -758,7 +765,6 @@ TRACE_EVENT(drv_ampdu_action, __field(u32, action) __field(u16, tid) __field(u16, ssn) - __field(int, ret) VIF_ENTRY ), @@ -766,15 +772,14 @@ TRACE_EVENT(drv_ampdu_action, LOCAL_ASSIGN; VIF_ASSIGN; STA_ASSIGN; - __entry->ret = ret; __entry->action = action; __entry->tid = tid; __entry->ssn = ssn ? *ssn : 0; ), TP_printk( - LOCAL_PR_FMT VIF_PR_FMT STA_PR_FMT " action:%d tid:%d ret:%d", - LOCAL_PR_ARG, VIF_PR_ARG, STA_PR_ARG, __entry->action, __entry->tid, __entry->ret + LOCAL_PR_FMT VIF_PR_FMT STA_PR_FMT " action:%d tid:%d", + LOCAL_PR_ARG, VIF_PR_ARG, STA_PR_ARG, __entry->action, __entry->tid ) ); -- cgit v1.1 From fbd2c8dcbc69616d2e15b8a269a86b3a05d45aea Mon Sep 17 00:00:00 2001 From: Teemu Paasikivi Date: Mon, 14 Jun 2010 12:55:31 +0300 Subject: mac80211: Set basic rates while joining ibss network This patch adds support to nl80211 and mac80211 to set basic rates when joining/creating ibss network. Original patch was posted by Johannes Berg on the linux-wireless posting list. Signed-off-by: Johannes Berg Signed-off-by: Teemu Paasikivi Signed-off-by: John W. Linville --- net/mac80211/ibss.c | 4 +++- net/mac80211/ieee80211_i.h | 2 ++ net/wireless/nl80211.c | 49 ++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 54 insertions(+), 1 deletion(-) (limited to 'net') diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index bfd72864..9f4e64e 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -172,6 +172,7 @@ static void __ieee80211_sta_join_ibss(struct ieee80211_sub_if_data *sdata, rcu_assign_pointer(ifibss->presp, skb); sdata->vif.bss_conf.beacon_int = beacon_int; + sdata->vif.bss_conf.basic_rates = basic_rates; bss_change = BSS_CHANGED_BEACON_INT; bss_change |= ieee80211_reset_erp_info(sdata); bss_change |= BSS_CHANGED_BSSID; @@ -529,7 +530,7 @@ static void ieee80211_sta_create_ibss(struct ieee80211_sub_if_data *sdata) sdata->drop_unencrypted = 0; __ieee80211_sta_join_ibss(sdata, bssid, sdata->vif.bss_conf.beacon_int, - ifibss->channel, 3, /* first two are basic */ + ifibss->channel, ifibss->basic_rates, capability, 0); } @@ -859,6 +860,7 @@ int ieee80211_ibss_join(struct ieee80211_sub_if_data *sdata, sdata->u.ibss.fixed_bssid = false; sdata->u.ibss.privacy = params->privacy; + sdata->u.ibss.basic_rates = params->basic_rates; sdata->vif.bss_conf.beacon_int = params->beacon_interval; diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 9d753a0..c3c2be3 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -387,6 +387,8 @@ struct ieee80211_if_ibss { unsigned long request; unsigned long last_scan_completed; + u32 basic_rates; + bool timer_running; bool fixed_bssid; diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index c65e67e..41529ac 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -3955,6 +3955,55 @@ static int nl80211_join_ibss(struct sk_buff *skb, struct genl_info *info) } } + if (info->attrs[NL80211_ATTR_BSS_BASIC_RATES]) { + u8 *rates = + nla_data(info->attrs[NL80211_ATTR_BSS_BASIC_RATES]); + int n_rates = + nla_len(info->attrs[NL80211_ATTR_BSS_BASIC_RATES]); + struct ieee80211_supported_band *sband = + wiphy->bands[ibss.channel->band]; + int i, j; + + if (n_rates == 0) { + err = -EINVAL; + goto out; + } + + for (i = 0; i < n_rates; i++) { + int rate = (rates[i] & 0x7f) * 5; + bool found = false; + + for (j = 0; j < sband->n_bitrates; j++) { + if (sband->bitrates[j].bitrate == rate) { + found = true; + ibss.basic_rates |= BIT(j); + break; + } + } + if (!found) { + err = -EINVAL; + goto out; + } + } + } else { + /* + * If no rates were explicitly configured, + * use the mandatory rate set for 11b or + * 11a for maximum compatibility. + */ + struct ieee80211_supported_band *sband = + wiphy->bands[ibss.channel->band]; + int j; + u32 flag = ibss.channel->band == IEEE80211_BAND_5GHZ ? + IEEE80211_RATE_MANDATORY_A : + IEEE80211_RATE_MANDATORY_B; + + for (j = 0; j < sband->n_bitrates; j++) { + if (sband->bitrates[j].flags & flag) + ibss.basic_rates |= BIT(j); + } + } + err = cfg80211_join_ibss(rdev, dev, &ibss, connkeys); out: -- cgit v1.1 From 392cfdb10dab6c7dfa5fed18d8a44d7453d42196 Mon Sep 17 00:00:00 2001 From: Teemu Paasikivi Date: Mon, 14 Jun 2010 12:55:32 +0300 Subject: mac80211: Set changed basic rates flag Add changed basic rates flag to bss_changed while joinig ibss network. This patch is split from the patch containing support for setting basic rates when creating ibss network. Original patch was posted by Johannes Berg on the linux-wireless posting list. Signed-off-by: Johannes Berg Signed-off-by: Teemu Paasikivi Signed-off-by: John W. Linville --- net/mac80211/ibss.c | 1 + 1 file changed, 1 insertion(+) (limited to 'net') diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index 9f4e64e..93481eb 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -178,6 +178,7 @@ static void __ieee80211_sta_join_ibss(struct ieee80211_sub_if_data *sdata, bss_change |= BSS_CHANGED_BSSID; bss_change |= BSS_CHANGED_BEACON; bss_change |= BSS_CHANGED_BEACON_ENABLED; + bss_change |= BSS_CHANGED_BASIC_RATES; bss_change |= BSS_CHANGED_IBSS; sdata->vif.bss_conf.ibss_joined = true; ieee80211_bss_info_change_notify(sdata, bss_change); -- cgit v1.1 From 5ea096c0c85e80335889539899af9a4717976e0b Mon Sep 17 00:00:00 2001 From: Teemu Paasikivi Date: Mon, 14 Jun 2010 12:55:33 +0300 Subject: mac80211: remove BSS from cfg80211 list when leaving IBSS Remove BSS from cfg80211 BSS list if we are only member in IBSS when leaving it. Signed-off-by: Teemu Paasikivi Acked-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/ibss.c | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) (limited to 'net') diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c index 93481eb..d4e84b2 100644 --- a/net/mac80211/ibss.c +++ b/net/mac80211/ibss.c @@ -917,6 +917,31 @@ int ieee80211_ibss_join(struct ieee80211_sub_if_data *sdata, int ieee80211_ibss_leave(struct ieee80211_sub_if_data *sdata) { struct sk_buff *skb; + struct ieee80211_if_ibss *ifibss = &sdata->u.ibss; + struct ieee80211_local *local = sdata->local; + struct cfg80211_bss *cbss; + u16 capability; + int active_ibss = 0; + + active_ibss = ieee80211_sta_active_ibss(sdata); + + if (!active_ibss && !is_zero_ether_addr(ifibss->bssid)) { + capability = WLAN_CAPABILITY_IBSS; + + if (ifibss->privacy) + capability |= WLAN_CAPABILITY_PRIVACY; + + cbss = cfg80211_get_bss(local->hw.wiphy, ifibss->channel, + ifibss->bssid, ifibss->ssid, + ifibss->ssid_len, WLAN_CAPABILITY_IBSS | + WLAN_CAPABILITY_PRIVACY, + capability); + + if (cbss) { + cfg80211_unlink_bss(local->hw.wiphy, cbss); + cfg80211_put_bss(cbss); + } + } del_timer_sync(&sdata->u.ibss.timer); clear_bit(IEEE80211_IBSS_REQ_RUN, &sdata->u.ibss.request); -- cgit v1.1 From 685429623f88d84f98bd5daffc3c427c408740d4 Mon Sep 17 00:00:00 2001 From: Juuso Oikarinen Date: Wed, 9 Jun 2010 13:43:26 +0300 Subject: mac80211: Fix circular locking dependency in ARP filter handling There is a circular locking dependency when configuring the hardware ARP filters on association, occurring when flushing the mac80211 workqueue. This is what happens: [ 92.026800] ======================================================= [ 92.030507] [ INFO: possible circular locking dependency detected ] [ 92.030507] 2.6.34-04781-g2b2c009 #85 [ 92.030507] ------------------------------------------------------- [ 92.030507] modprobe/5225 is trying to acquire lock: [ 92.030507] ((wiphy_name(local->hw.wiphy))){+.+.+.}, at: [] flush_workq ueue+0x0/0xb0 [ 92.030507] [ 92.030507] but task is already holding lock: [ 92.030507] (rtnl_mutex){+.+.+.}, at: [] rtnl_lock+0x12/0x20 [ 92.030507] [ 92.030507] which lock already depends on the new lock. [ 92.030507] [ 92.030507] [ 92.030507] the existing dependency chain (in reverse order) is: [ 92.030507] [ 92.030507] -> #2 (rtnl_mutex){+.+.+.}: [ 92.030507] [] lock_acquire+0xdb/0x110 [ 92.030507] [] mutex_lock_nested+0x44/0x300 [ 92.030507] [] rtnl_lock+0x12/0x20 [ 92.030507] [] ieee80211_assoc_done+0x6c/0xe0 [mac80211] [ 92.030507] [] ieee80211_work_work+0x31d/0x1280 [mac80211] [ 92.030507] -> #1 ((&local->work_work)){+.+.+.}: [ 92.030507] [] lock_acquire+0xdb/0x110 [ 92.030507] [] worker_thread+0x22a/0x370 [ 92.030507] [] kthread+0x96/0xb0 [ 92.030507] [] kernel_thread_helper+0x4/0x10 [ 92.030507] [ 92.030507] -> #0 ((wiphy_name(local->hw.wiphy))){+.+.+.}: [ 92.030507] [] __lock_acquire+0x1c0c/0x1d50 [ 92.030507] [] lock_acquire+0xdb/0x110 [ 92.030507] [] flush_workqueue+0x4e/0xb0 [ 92.030507] [] ieee80211_stop_device+0x2b/0xb0 [mac80211] [ 92.030507] [] ieee80211_stop+0x3e5/0x680 [mac80211] The locking in this case is quite complex. Fix the problem by rewriting the way the hardware ARP filter list is handled - i.e. make a copy of the address list to the bss_conf struct, and provide that list to the hardware driver when needed. The current patch will enable filtering also in promiscuous mode. This may need to be changed in the future. Reported-by: Reinette Chatre Signed-off-by: Juuso Oikarinen Signed-off-by: John W. Linville --- net/mac80211/driver-ops.h | 17 -------------- net/mac80211/driver-trace.h | 22 ------------------ net/mac80211/ieee80211_i.h | 2 ++ net/mac80211/iface.c | 3 +++ net/mac80211/main.c | 54 ++++++++++++++++++++++++++++----------------- net/mac80211/mlme.c | 34 +++++++++++++++------------- 6 files changed, 58 insertions(+), 74 deletions(-) (limited to 'net') diff --git a/net/mac80211/driver-ops.h b/net/mac80211/driver-ops.h index 965d64f..c333173 100644 --- a/net/mac80211/driver-ops.h +++ b/net/mac80211/driver-ops.h @@ -89,23 +89,6 @@ static inline void drv_bss_info_changed(struct ieee80211_local *local, trace_drv_return_void(local); } -struct in_ifaddr; -static inline int drv_configure_arp_filter(struct ieee80211_local *local, - struct ieee80211_vif *vif, - struct in_ifaddr *ifa_list) -{ - int ret = 0; - - might_sleep(); - - trace_drv_configure_arp_filter(local, vif_to_sdata(vif)); - if (local->ops->configure_arp_filter) - ret = local->ops->configure_arp_filter(&local->hw, vif, - ifa_list); - trace_drv_return_int(local, ret); - return ret; -} - static inline u64 drv_prepare_multicast(struct ieee80211_local *local, struct netdev_hw_addr_list *mc_list) { diff --git a/net/mac80211/driver-trace.h b/net/mac80211/driver-trace.h index 06444ea..8da31ca 100644 --- a/net/mac80211/driver-trace.h +++ b/net/mac80211/driver-trace.h @@ -251,28 +251,6 @@ TRACE_EVENT(drv_bss_info_changed, ) ); -TRACE_EVENT(drv_configure_arp_filter, - TP_PROTO(struct ieee80211_local *local, - struct ieee80211_sub_if_data *sdata), - - TP_ARGS(local, sdata), - - TP_STRUCT__entry( - LOCAL_ENTRY - VIF_ENTRY - ), - - TP_fast_assign( - LOCAL_ASSIGN; - VIF_ASSIGN; - ), - - TP_printk( - VIF_PR_FMT LOCAL_PR_FMT, - VIF_PR_ARG, LOCAL_PR_ARG - ) -); - TRACE_EVENT(drv_prepare_multicast, TP_PROTO(struct ieee80211_local *local, int mc_count), diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index c3c2be3..9b3c3f9 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -514,6 +514,8 @@ struct ieee80211_sub_if_data { struct work_struct work; struct sk_buff_head skb_queue; + bool arp_filter_state; + /* * AP this belongs to: self in AP mode and * corresponding AP in VLAN mode, NULL for diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 490be2f..910729f 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -1076,6 +1076,9 @@ int ieee80211_if_add(struct ieee80211_local *local, const char *name, sdata->wdev.wiphy = local->hw.wiphy; sdata->local = local; sdata->dev = ndev; +#ifdef CONFIG_INET + sdata->arp_filter_state = true; +#endif for (i = 0; i < IEEE80211_FRAGMENT_MAX; i++) skb_queue_head_init(&sdata->fragments[i].skb_list); diff --git a/net/mac80211/main.c b/net/mac80211/main.c index c2e46e8..a1bf46c 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include @@ -317,23 +318,6 @@ static void ieee80211_recalc_smps_work(struct work_struct *work) } #ifdef CONFIG_INET -int ieee80211_set_arp_filter(struct ieee80211_sub_if_data *sdata) -{ - struct in_device *idev; - int ret = 0; - - BUG_ON(!sdata); - ASSERT_RTNL(); - - idev = sdata->dev->ip_ptr; - if (!idev) - return 0; - - ret = drv_configure_arp_filter(sdata->local, &sdata->vif, - idev->ifa_list); - return ret; -} - static int ieee80211_ifa_changed(struct notifier_block *nb, unsigned long data, void *arg) { @@ -343,8 +327,11 @@ static int ieee80211_ifa_changed(struct notifier_block *nb, ifa_notifier); struct net_device *ndev = ifa->ifa_dev->dev; struct wireless_dev *wdev = ndev->ieee80211_ptr; + struct in_device *idev; struct ieee80211_sub_if_data *sdata; + struct ieee80211_bss_conf *bss_conf; struct ieee80211_if_managed *ifmgd; + int c = 0; if (!netif_running(ndev)) return NOTIFY_DONE; @@ -356,17 +343,44 @@ static int ieee80211_ifa_changed(struct notifier_block *nb, if (wdev->wiphy != local->hw.wiphy) return NOTIFY_DONE; - /* We are concerned about IP addresses only when associated */ sdata = IEEE80211_DEV_TO_SUB_IF(ndev); + bss_conf = &sdata->vif.bss_conf; /* ARP filtering is only supported in managed mode */ if (sdata->vif.type != NL80211_IFTYPE_STATION) return NOTIFY_DONE; + idev = sdata->dev->ip_ptr; + if (!idev) + return NOTIFY_DONE; + ifmgd = &sdata->u.mgd; mutex_lock(&ifmgd->mtx); - if (ifmgd->associated) - ieee80211_set_arp_filter(sdata); + + /* Copy the addresses to the bss_conf list */ + ifa = idev->ifa_list; + while (c < IEEE80211_BSS_ARP_ADDR_LIST_LEN && ifa) { + bss_conf->arp_addr_list[c] = ifa->ifa_address; + ifa = ifa->ifa_next; + c++; + } + + /* If not all addresses fit the list, disable filtering */ + if (ifa) { + sdata->arp_filter_state = false; + c = 0; + } else { + sdata->arp_filter_state = true; + } + bss_conf->arp_addr_cnt = c; + + /* Configure driver only if associated */ + if (ifmgd->associated) { + bss_conf->arp_filter_enabled = sdata->arp_filter_state; + ieee80211_bss_info_change_notify(sdata, + BSS_CHANGED_ARP_FILTER); + } + mutex_unlock(&ifmgd->mtx); return NOTIFY_DONE; diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 583b346..74479c2 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -806,11 +806,12 @@ static void ieee80211_set_associated(struct ieee80211_sub_if_data *sdata, { struct ieee80211_bss *bss = (void *)cbss->priv; struct ieee80211_local *local = sdata->local; + struct ieee80211_bss_conf *bss_conf = &sdata->vif.bss_conf; bss_info_changed |= BSS_CHANGED_ASSOC; /* set timing information */ - sdata->vif.bss_conf.beacon_int = cbss->beacon_interval; - sdata->vif.bss_conf.timestamp = cbss->tsf; + bss_conf->beacon_int = cbss->beacon_interval; + bss_conf->timestamp = cbss->tsf; bss_info_changed |= BSS_CHANGED_BEACON_INT; bss_info_changed |= ieee80211_handle_bss_capability(sdata, @@ -835,7 +836,7 @@ static void ieee80211_set_associated(struct ieee80211_sub_if_data *sdata, ieee80211_led_assoc(local, 1); - sdata->vif.bss_conf.assoc = 1; + bss_conf->assoc = 1; /* * For now just always ask the driver to update the basic rateset * when we have associated, we aren't checking whether it actually @@ -848,9 +849,15 @@ static void ieee80211_set_associated(struct ieee80211_sub_if_data *sdata, /* Tell the driver to monitor connection quality (if supported) */ if ((local->hw.flags & IEEE80211_HW_SUPPORTS_CQM_RSSI) && - sdata->vif.bss_conf.cqm_rssi_thold) + bss_conf->cqm_rssi_thold) bss_info_changed |= BSS_CHANGED_CQM; + /* Enable ARP filtering */ + if (bss_conf->arp_filter_enabled != sdata->arp_filter_state) { + bss_conf->arp_filter_enabled = sdata->arp_filter_state; + bss_info_changed |= BSS_CHANGED_ARP_FILTER; + } + ieee80211_bss_info_change_notify(sdata, bss_info_changed); mutex_lock(&local->iflist_mtx); @@ -932,6 +939,12 @@ static void ieee80211_set_disassoc(struct ieee80211_sub_if_data *sdata, ieee80211_hw_config(local, config_changed); + /* Disable ARP filtering */ + if (sdata->vif.bss_conf.arp_filter_enabled) { + sdata->vif.bss_conf.arp_filter_enabled = false; + changed |= BSS_CHANGED_ARP_FILTER; + } + /* The BSSID (not really interesting) and HT changed */ changed |= BSS_CHANGED_BSSID | BSS_CHANGED_HT; ieee80211_bss_info_change_notify(sdata, changed); @@ -2018,18 +2031,9 @@ static enum work_done_result ieee80211_assoc_done(struct ieee80211_work *wk, cfg80211_send_assoc_timeout(wk->sdata->dev, wk->filter_ta); return WORK_DONE_DESTROY; - } else { - mutex_unlock(&wk->sdata->u.mgd.mtx); -#ifdef CONFIG_INET - /* - * configure ARP filter IP addresses to the driver, - * intentionally outside the mgd mutex. - */ - rtnl_lock(); - ieee80211_set_arp_filter(wk->sdata); - rtnl_unlock(); -#endif } + + mutex_unlock(&wk->sdata->u.mgd.mtx); } cfg80211_send_rx_assoc(wk->sdata->dev, skb->data, skb->len); -- cgit v1.1 From f6bc7d9e4760324258ad5f5d147e79db8442842e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 14 Jun 2010 04:39:27 +0000 Subject: ipv6: avoid two atomics in ipv6_rthdr_rcv() Use __in6_dev_get() instead of in6_dev_get()/in6_dev_put() Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv6/exthdrs.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) (limited to 'net') diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c index 853a633..262f105 100644 --- a/net/ipv6/exthdrs.c +++ b/net/ipv6/exthdrs.c @@ -312,6 +312,7 @@ static int ipv6_destopt_rcv(struct sk_buff *skb) Routing header. ********************************/ +/* called with rcu_read_lock() */ static int ipv6_rthdr_rcv(struct sk_buff *skb) { struct inet6_skb_parm *opt = IP6CB(skb); @@ -324,12 +325,9 @@ static int ipv6_rthdr_rcv(struct sk_buff *skb) struct net *net = dev_net(skb->dev); int accept_source_route = net->ipv6.devconf_all->accept_source_route; - idev = in6_dev_get(skb->dev); - if (idev) { - if (accept_source_route > idev->cnf.accept_source_route) - accept_source_route = idev->cnf.accept_source_route; - in6_dev_put(idev); - } + idev = __in6_dev_get(skb->dev); + if (idev && accept_source_route > idev->cnf.accept_source_route) + accept_source_route = idev->cnf.accept_source_route; if (!pskb_may_pull(skb, skb_transport_offset(skb) + 8) || !pskb_may_pull(skb, (skb_transport_offset(skb) + -- cgit v1.1 From c68f24cc354050415c5ea543cd19ea5424463a2f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 14 Jun 2010 04:46:20 +0000 Subject: ipv6: RCU changes in ipv6_get_mtu() and ip6_dst_hoplimit() Use RCU to avoid atomic ops on idev refcnt in ipv6_get_mtu() and ip6_dst_hoplimit() Signed-off-by: Eric Dumazet Acked-by: YOSHIFUJI Hideaki Signed-off-by: David S. Miller --- net/ipv6/route.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) (limited to 'net') diff --git a/net/ipv6/route.c b/net/ipv6/route.c index f770285..8f2d040 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1084,11 +1084,11 @@ static int ipv6_get_mtu(struct net_device *dev) int mtu = IPV6_MIN_MTU; struct inet6_dev *idev; - idev = in6_dev_get(dev); - if (idev) { + rcu_read_lock(); + idev = __in6_dev_get(dev); + if (idev) mtu = idev->cnf.mtu6; - in6_dev_put(idev); - } + rcu_read_unlock(); return mtu; } @@ -1097,12 +1097,15 @@ int ip6_dst_hoplimit(struct dst_entry *dst) int hoplimit = dst_metric(dst, RTAX_HOPLIMIT); if (hoplimit < 0) { struct net_device *dev = dst->dev; - struct inet6_dev *idev = in6_dev_get(dev); - if (idev) { + struct inet6_dev *idev; + + rcu_read_lock(); + idev = __in6_dev_get(dev); + if (idev) hoplimit = idev->cnf.hop_limit; - in6_dev_put(idev); - } else + else hoplimit = dev_net(dev)->ipv6.devconf_all->hop_limit; + rcu_read_unlock(); } return hoplimit; } -- cgit v1.1 From d6cc1d642de9284cb26488ea390d915b50ee2504 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 14 Jun 2010 19:35:21 +0000 Subject: inetpeer: various changes Try to reduce cache line contentions in peer management, to reduce IP defragmentation overhead. - peer_fake_node is marked 'const' to make sure its not modified. (tested with CONFIG_DEBUG_RODATA=y) - Group variables in two structures to reduce number of dirtied cache lines. One named "peers" for avl tree root, its number of entries, and associated lock. (candidate for RCU conversion) - A second one named "unused_peers" for unused list and its lock - Add a !list_empty() test in unlink_from_unused() to avoid taking lock when entry is not unused. - Use atomic_dec_and_lock() in inet_putpeer() to avoid taking lock in some cases. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/inetpeer.c | 94 +++++++++++++++++++++++++++++++---------------------- 1 file changed, 56 insertions(+), 38 deletions(-) (limited to 'net') diff --git a/net/ipv4/inetpeer.c b/net/ipv4/inetpeer.c index 6bcfe52..035673fd 100644 --- a/net/ipv4/inetpeer.c +++ b/net/ipv4/inetpeer.c @@ -70,17 +70,25 @@ static struct kmem_cache *peer_cachep __read_mostly; #define node_height(x) x->avl_height -static struct inet_peer peer_fake_node = { - .avl_left = &peer_fake_node, - .avl_right = &peer_fake_node, + +#define peer_avl_empty ((struct inet_peer *)&peer_fake_node) +static const struct inet_peer peer_fake_node = { + .avl_left = peer_avl_empty, + .avl_right = peer_avl_empty, .avl_height = 0 }; -#define peer_avl_empty (&peer_fake_node) -static struct inet_peer *peer_root = peer_avl_empty; -static DEFINE_RWLOCK(peer_pool_lock); + +static struct { + struct inet_peer *root; + rwlock_t lock; + int total; +} peers = { + .root = peer_avl_empty, + .lock = __RW_LOCK_UNLOCKED(peers.lock), + .total = 0, +}; #define PEER_MAXDEPTH 40 /* sufficient for about 2^27 nodes */ -static int peer_total; /* Exported for sysctl_net_ipv4. */ int inet_peer_threshold __read_mostly = 65536 + 128; /* start to throw entries more * aggressively at this stage */ @@ -89,8 +97,13 @@ int inet_peer_maxttl __read_mostly = 10 * 60 * HZ; /* usual time to live: 10 min int inet_peer_gc_mintime __read_mostly = 10 * HZ; int inet_peer_gc_maxtime __read_mostly = 120 * HZ; -static LIST_HEAD(unused_peers); -static DEFINE_SPINLOCK(inet_peer_unused_lock); +static struct { + struct list_head list; + spinlock_t lock; +} unused_peers = { + .list = LIST_HEAD_INIT(unused_peers.list), + .lock = __SPIN_LOCK_UNLOCKED(unused_peers.lock), +}; static void peer_check_expire(unsigned long dummy); static DEFINE_TIMER(peer_periodic_timer, peer_check_expire, 0, 0); @@ -131,9 +144,11 @@ void __init inet_initpeers(void) /* Called with or without local BH being disabled. */ static void unlink_from_unused(struct inet_peer *p) { - spin_lock_bh(&inet_peer_unused_lock); - list_del_init(&p->unused); - spin_unlock_bh(&inet_peer_unused_lock); + if (!list_empty(&p->unused)) { + spin_lock_bh(&unused_peers.lock); + list_del_init(&p->unused); + spin_unlock_bh(&unused_peers.lock); + } } /* @@ -146,9 +161,9 @@ static void unlink_from_unused(struct inet_peer *p) struct inet_peer *u, **v; \ if (_stack != NULL) { \ stackptr = _stack; \ - *stackptr++ = &peer_root; \ + *stackptr++ = &peers.root; \ } \ - for (u = peer_root; u != peer_avl_empty; ) { \ + for (u = peers.root; u != peer_avl_empty; ) { \ if (_daddr == u->v4daddr) \ break; \ if ((__force __u32)_daddr < (__force __u32)u->v4daddr) \ @@ -262,7 +277,7 @@ do { \ n->avl_right = peer_avl_empty; \ **--stackptr = n; \ peer_avl_rebalance(stack, stackptr); \ -} while(0) +} while (0) /* May be called with local BH enabled. */ static void unlink_from_pool(struct inet_peer *p) @@ -271,7 +286,7 @@ static void unlink_from_pool(struct inet_peer *p) do_free = 0; - write_lock_bh(&peer_pool_lock); + write_lock_bh(&peers.lock); /* Check the reference counter. It was artificially incremented by 1 * in cleanup() function to prevent sudden disappearing. If the * reference count is still 1 then the node is referenced only as `p' @@ -303,10 +318,10 @@ static void unlink_from_pool(struct inet_peer *p) delp[1] = &t->avl_left; /* was &p->avl_left */ } peer_avl_rebalance(stack, stackptr); - peer_total--; + peers.total--; do_free = 1; } - write_unlock_bh(&peer_pool_lock); + write_unlock_bh(&peers.lock); if (do_free) kmem_cache_free(peer_cachep, p); @@ -326,16 +341,16 @@ static int cleanup_once(unsigned long ttl) struct inet_peer *p = NULL; /* Remove the first entry from the list of unused nodes. */ - spin_lock_bh(&inet_peer_unused_lock); - if (!list_empty(&unused_peers)) { + spin_lock_bh(&unused_peers.lock); + if (!list_empty(&unused_peers.list)) { __u32 delta; - p = list_first_entry(&unused_peers, struct inet_peer, unused); + p = list_first_entry(&unused_peers.list, struct inet_peer, unused); delta = (__u32)jiffies - p->dtime; if (delta < ttl) { /* Do not prune fresh entries. */ - spin_unlock_bh(&inet_peer_unused_lock); + spin_unlock_bh(&unused_peers.lock); return -1; } @@ -345,7 +360,7 @@ static int cleanup_once(unsigned long ttl) * before unlink_from_pool() call. */ atomic_inc(&p->refcnt); } - spin_unlock_bh(&inet_peer_unused_lock); + spin_unlock_bh(&unused_peers.lock); if (p == NULL) /* It means that the total number of USED entries has @@ -364,11 +379,11 @@ struct inet_peer *inet_getpeer(__be32 daddr, int create) struct inet_peer **stack[PEER_MAXDEPTH], ***stackptr; /* Look up for the address quickly. */ - read_lock_bh(&peer_pool_lock); + read_lock_bh(&peers.lock); p = lookup(daddr, NULL); if (p != peer_avl_empty) atomic_inc(&p->refcnt); - read_unlock_bh(&peer_pool_lock); + read_unlock_bh(&peers.lock); if (p != peer_avl_empty) { /* The existing node has been found. */ @@ -390,7 +405,7 @@ struct inet_peer *inet_getpeer(__be32 daddr, int create) atomic_set(&n->ip_id_count, secure_ip_id(daddr)); n->tcp_ts_stamp = 0; - write_lock_bh(&peer_pool_lock); + write_lock_bh(&peers.lock); /* Check if an entry has suddenly appeared. */ p = lookup(daddr, stack); if (p != peer_avl_empty) @@ -399,10 +414,10 @@ struct inet_peer *inet_getpeer(__be32 daddr, int create) /* Link the node. */ link_to_pool(n); INIT_LIST_HEAD(&n->unused); - peer_total++; - write_unlock_bh(&peer_pool_lock); + peers.total++; + write_unlock_bh(&peers.lock); - if (peer_total >= inet_peer_threshold) + if (peers.total >= inet_peer_threshold) /* Remove one less-recently-used entry. */ cleanup_once(0); @@ -411,7 +426,7 @@ struct inet_peer *inet_getpeer(__be32 daddr, int create) out_free: /* The appropriate node is already in the pool. */ atomic_inc(&p->refcnt); - write_unlock_bh(&peer_pool_lock); + write_unlock_bh(&peers.lock); /* Remove the entry from unused list if it was there. */ unlink_from_unused(p); /* Free preallocated the preallocated node. */ @@ -425,12 +440,12 @@ static void peer_check_expire(unsigned long dummy) unsigned long now = jiffies; int ttl; - if (peer_total >= inet_peer_threshold) + if (peers.total >= inet_peer_threshold) ttl = inet_peer_minttl; else ttl = inet_peer_maxttl - (inet_peer_maxttl - inet_peer_minttl) / HZ * - peer_total / inet_peer_threshold * HZ; + peers.total / inet_peer_threshold * HZ; while (!cleanup_once(ttl)) { if (jiffies != now) break; @@ -439,22 +454,25 @@ static void peer_check_expire(unsigned long dummy) /* Trigger the timer after inet_peer_gc_mintime .. inet_peer_gc_maxtime * interval depending on the total number of entries (more entries, * less interval). */ - if (peer_total >= inet_peer_threshold) + if (peers.total >= inet_peer_threshold) peer_periodic_timer.expires = jiffies + inet_peer_gc_mintime; else peer_periodic_timer.expires = jiffies + inet_peer_gc_maxtime - (inet_peer_gc_maxtime - inet_peer_gc_mintime) / HZ * - peer_total / inet_peer_threshold * HZ; + peers.total / inet_peer_threshold * HZ; add_timer(&peer_periodic_timer); } void inet_putpeer(struct inet_peer *p) { - spin_lock_bh(&inet_peer_unused_lock); - if (atomic_dec_and_test(&p->refcnt)) { - list_add_tail(&p->unused, &unused_peers); + local_bh_disable(); + + if (atomic_dec_and_lock(&p->refcnt, &unused_peers.lock)) { + list_add_tail(&p->unused, &unused_peers.list); p->dtime = (__u32)jiffies; + spin_unlock(&unused_peers.lock); } - spin_unlock_bh(&inet_peer_unused_lock); + + local_bh_enable(); } -- cgit v1.1 From d73f33b168831e53972fbf7c85db87950a41436c Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 15 Jun 2010 13:08:51 +0200 Subject: netfilter: CLUSTERIP: RCU conversion - clusterip_lock becomes a spinlock - lockless lookups - kfree() deferred after RCU grace period - rcu_barrier_bh() inserted in clusterip_tg_exit() v2) - As Patrick pointed out, we use atomic_inc_not_zero() in clusterip_config_find_get(). - list_add_rcu() and list_del_rcu() variants are used. - atomic_dec_and_lock() used in clusterip_config_entry_put() Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/ipv4/netfilter/ipt_CLUSTERIP.c | 48 +++++++++++++++++++++++--------------- 1 file changed, 29 insertions(+), 19 deletions(-) (limited to 'net') diff --git a/net/ipv4/netfilter/ipt_CLUSTERIP.c b/net/ipv4/netfilter/ipt_CLUSTERIP.c index f91c94b..64d0875 100644 --- a/net/ipv4/netfilter/ipt_CLUSTERIP.c +++ b/net/ipv4/netfilter/ipt_CLUSTERIP.c @@ -53,12 +53,13 @@ struct clusterip_config { #endif enum clusterip_hashmode hash_mode; /* which hashing mode */ u_int32_t hash_initval; /* hash initialization */ + struct rcu_head rcu; }; static LIST_HEAD(clusterip_configs); /* clusterip_lock protects the clusterip_configs list */ -static DEFINE_RWLOCK(clusterip_lock); +static DEFINE_SPINLOCK(clusterip_lock); #ifdef CONFIG_PROC_FS static const struct file_operations clusterip_proc_fops; @@ -71,11 +72,17 @@ clusterip_config_get(struct clusterip_config *c) atomic_inc(&c->refcount); } + +static void clusterip_config_rcu_free(struct rcu_head *head) +{ + kfree(container_of(head, struct clusterip_config, rcu)); +} + static inline void clusterip_config_put(struct clusterip_config *c) { if (atomic_dec_and_test(&c->refcount)) - kfree(c); + call_rcu_bh(&c->rcu, clusterip_config_rcu_free); } /* decrease the count of entries using/referencing this config. If last @@ -84,10 +91,11 @@ clusterip_config_put(struct clusterip_config *c) static inline void clusterip_config_entry_put(struct clusterip_config *c) { - write_lock_bh(&clusterip_lock); - if (atomic_dec_and_test(&c->entries)) { - list_del(&c->list); - write_unlock_bh(&clusterip_lock); + local_bh_disable(); + if (atomic_dec_and_lock(&c->entries, &clusterip_lock)) { + list_del_rcu(&c->list); + spin_unlock(&clusterip_lock); + local_bh_enable(); dev_mc_del(c->dev, c->clustermac); dev_put(c->dev); @@ -100,7 +108,7 @@ clusterip_config_entry_put(struct clusterip_config *c) #endif return; } - write_unlock_bh(&clusterip_lock); + local_bh_enable(); } static struct clusterip_config * @@ -108,7 +116,7 @@ __clusterip_config_find(__be32 clusterip) { struct clusterip_config *c; - list_for_each_entry(c, &clusterip_configs, list) { + list_for_each_entry_rcu(c, &clusterip_configs, list) { if (c->clusterip == clusterip) return c; } @@ -121,16 +129,15 @@ clusterip_config_find_get(__be32 clusterip, int entry) { struct clusterip_config *c; - read_lock_bh(&clusterip_lock); + rcu_read_lock_bh(); c = __clusterip_config_find(clusterip); - if (!c) { - read_unlock_bh(&clusterip_lock); - return NULL; + if (c) { + if (unlikely(!atomic_inc_not_zero(&c->refcount))) + c = NULL; + else if (entry) + atomic_inc(&c->entries); } - atomic_inc(&c->refcount); - if (entry) - atomic_inc(&c->entries); - read_unlock_bh(&clusterip_lock); + rcu_read_unlock_bh(); return c; } @@ -181,9 +188,9 @@ clusterip_config_init(const struct ipt_clusterip_tgt_info *i, __be32 ip, } #endif - write_lock_bh(&clusterip_lock); - list_add(&c->list, &clusterip_configs); - write_unlock_bh(&clusterip_lock); + spin_lock_bh(&clusterip_lock); + list_add_rcu(&c->list, &clusterip_configs); + spin_unlock_bh(&clusterip_lock); return c; } @@ -733,6 +740,9 @@ static void __exit clusterip_tg_exit(void) #endif nf_unregister_hook(&cip_arp_ops); xt_unregister_target(&clusterip_tg_reg); + + /* Wait for completion of call_rcu_bh()'s (clusterip_config_rcu_free) */ + rcu_barrier_bh(); } module_init(clusterip_tg_init); -- cgit v1.1 From 0902b469bd25065aa0688c3cee6f11744c817e7c Mon Sep 17 00:00:00 2001 From: Luciano Coelho Date: Tue, 15 Jun 2010 15:04:00 +0200 Subject: netfilter: xtables: idletimer target implementation This patch implements an idletimer Xtables target that can be used to identify when interfaces have been idle for a certain period of time. Timers are identified by labels and are created when a rule is set with a new label. The rules also take a timeout value (in seconds) as an option. If more than one rule uses the same timer label, the timer will be restarted whenever any of the rules get a hit. One entry for each timer is created in sysfs. This attribute contains the timer remaining for the timer to expire. The attributes are located under the xt_idletimer class: /sys/class/xt_idletimer/timers/