From 358b1361bed42f4e6cbf8956a73aebf193957d4a Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Thu, 21 Apr 2011 10:55:07 +0200 Subject: [PATCH 01/28] netfilter: ctnetlink: fix timestamp support for new conntracks This patch fixes the missing initialization of the start time if the timestamp support is enabled. libnetfilter_conntrack/utils# conntrack -E & libnetfilter_conntrack/utils# ./conntrack_create tcp 6 109 ESTABLISHED src=1.1.1.1 dst=2.2.2.2 sport=1025 dport=21 packets=0 bytes=0 [UNREPLIED] src=2.2.2.2 dst=1.1.1.1 sport=21 dport=1025 packets=0 bytes=0 mark=0 delta-time=1303296401 use=2 Signed-off-by: Pablo Neira Ayuso Signed-off-by: Patrick McHardy --- net/netfilter/nf_conntrack_netlink.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index 30bf8a167fc8..482e90c61850 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -1334,6 +1334,7 @@ ctnetlink_create_conntrack(struct net *net, u16 zone, struct nf_conn *ct; int err = -EINVAL; struct nf_conntrack_helper *helper; + struct nf_conn_tstamp *tstamp; ct = nf_conntrack_alloc(net, zone, otuple, rtuple, GFP_ATOMIC); if (IS_ERR(ct)) @@ -1451,6 +1452,9 @@ ctnetlink_create_conntrack(struct net *net, u16 zone, __set_bit(IPS_EXPECTED_BIT, &ct->status); ct->master = master_ct; } + tstamp = nf_conn_tstamp_find(ct); + if (tstamp) + tstamp->start = ktime_to_ns(ktime_get_real()); add_timer(&ct->timeout); nf_conntrack_hash_insert(ct); From 954d70388307eb1ccb328d9f960657022a316243 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 21 Apr 2011 10:57:21 +0200 Subject: [PATCH 02/28] netfilter: fix ebtables compat support commit 255d0dc34068a976 (netfilter: x_table: speedup compat operations) made ebtables not working anymore. 1) xt_compat_calc_jump() is not an exact match lookup 2) compat_table_info() has a typo in xt_compat_init_offsets() call 3) compat_do_replace() misses a xt_compat_init_offsets() call Reported-by: dann frazier Signed-off-by: Eric Dumazet Signed-off-by: Patrick McHardy --- net/bridge/netfilter/ebtables.c | 3 ++- net/netfilter/x_tables.c | 4 ++-- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c index 893669caa8de..9707079bc40a 100644 --- a/net/bridge/netfilter/ebtables.c +++ b/net/bridge/netfilter/ebtables.c @@ -1766,7 +1766,7 @@ static int compat_table_info(const struct ebt_table_info *info, newinfo->entries_size = size; - xt_compat_init_offsets(AF_INET, info->nentries); + xt_compat_init_offsets(NFPROTO_BRIDGE, info->nentries); return EBT_ENTRY_ITERATE(entries, size, compat_calc_entry, info, entries, newinfo); } @@ -2240,6 +2240,7 @@ static int compat_do_replace(struct net *net, void __user *user, xt_compat_lock(NFPROTO_BRIDGE); + xt_compat_init_offsets(NFPROTO_BRIDGE, tmp.nentries); ret = compat_copy_entries(entries_tmp, tmp.entries_size, &state); if (ret < 0) goto out_unlock; diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c index a9adf4c6b299..8a025a585d2f 100644 --- a/net/netfilter/x_tables.c +++ b/net/netfilter/x_tables.c @@ -455,6 +455,7 @@ void xt_compat_flush_offsets(u_int8_t af) vfree(xt[af].compat_tab); xt[af].compat_tab = NULL; xt[af].number = 0; + xt[af].cur = 0; } } EXPORT_SYMBOL_GPL(xt_compat_flush_offsets); @@ -473,8 +474,7 @@ int xt_compat_calc_jump(u_int8_t af, unsigned int offset) else return mid ? tmp[mid - 1].delta : 0; } - WARN_ON_ONCE(1); - return 0; + return left ? tmp[left - 1].delta : 0; } EXPORT_SYMBOL_GPL(xt_compat_calc_jump); From e31b2b228439eee23b8ccc186d3c0e77bd69123b Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 21 Apr 2011 10:58:25 +0200 Subject: [PATCH 03/28] netfilter: ebtables: only call xt_compat_add_offset once per rule The optimizations in commit 255d0dc34068a976 (netfilter: x_table: speedup compat operations) assume that xt_compat_add_offset is called once per rule. ebtables however called it for each match/target found in a rule. The match/watcher/target parser already returns the needed delta, so it is sufficient to move the xt_compat_add_offset call to a more reasonable location. While at it, also get rid of the unused COMPAT iterator macros. Signed-off-by: Florian Westphal Signed-off-by: Patrick McHardy --- net/bridge/netfilter/ebtables.c | 61 +++++---------------------------- 1 file changed, 9 insertions(+), 52 deletions(-) diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c index 9707079bc40a..1a92b369c820 100644 --- a/net/bridge/netfilter/ebtables.c +++ b/net/bridge/netfilter/ebtables.c @@ -1882,7 +1882,7 @@ static int compat_mtw_from_user(struct compat_ebt_entry_mwt *mwt, struct xt_match *match; struct xt_target *wt; void *dst = NULL; - int off, pad = 0, ret = 0; + int off, pad = 0; unsigned int size_kern, entry_offset, match_size = mwt->match_size; strlcpy(name, mwt->u.name, sizeof(name)); @@ -1935,13 +1935,6 @@ static int compat_mtw_from_user(struct compat_ebt_entry_mwt *mwt, break; } - if (!dst) { - ret = xt_compat_add_offset(NFPROTO_BRIDGE, entry_offset, - off + ebt_compat_entry_padsize()); - if (ret < 0) - return ret; - } - state->buf_kern_offset += match_size + off; state->buf_user_offset += match_size; pad = XT_ALIGN(size_kern) - size_kern; @@ -2016,50 +2009,6 @@ static int ebt_size_mwt(struct compat_ebt_entry_mwt *match32, return growth; } -#define EBT_COMPAT_WATCHER_ITERATE(e, fn, args...) \ -({ \ - unsigned int __i; \ - int __ret = 0; \ - struct compat_ebt_entry_mwt *__watcher; \ - \ - for (__i = e->watchers_offset; \ - __i < (e)->target_offset; \ - __i += __watcher->watcher_size + \ - sizeof(struct compat_ebt_entry_mwt)) { \ - __watcher = (void *)(e) + __i; \ - __ret = fn(__watcher , ## args); \ - if (__ret != 0) \ - break; \ - } \ - if (__ret == 0) { \ - if (__i != (e)->target_offset) \ - __ret = -EINVAL; \ - } \ - __ret; \ -}) - -#define EBT_COMPAT_MATCH_ITERATE(e, fn, args...) \ -({ \ - unsigned int __i; \ - int __ret = 0; \ - struct compat_ebt_entry_mwt *__match; \ - \ - for (__i = sizeof(struct ebt_entry); \ - __i < (e)->watchers_offset; \ - __i += __match->match_size + \ - sizeof(struct compat_ebt_entry_mwt)) { \ - __match = (void *)(e) + __i; \ - __ret = fn(__match , ## args); \ - if (__ret != 0) \ - break; \ - } \ - if (__ret == 0) { \ - if (__i != (e)->watchers_offset) \ - __ret = -EINVAL; \ - } \ - __ret; \ -}) - /* called for all ebt_entry structures. */ static int size_entry_mwt(struct ebt_entry *entry, const unsigned char *base, unsigned int *total, @@ -2132,6 +2081,14 @@ static int size_entry_mwt(struct ebt_entry *entry, const unsigned char *base, } } + if (state->buf_kern_start == NULL) { + unsigned int offset = buf_start - (char *) base; + + ret = xt_compat_add_offset(NFPROTO_BRIDGE, offset, new_offset); + if (ret < 0) + return ret; + } + startoff = state->buf_user_offset - startoff; BUG_ON(*total < startoff); From 421eab4cf318ce47c1daae391e54d1b427c6687b Mon Sep 17 00:00:00 2001 From: Hans Schillstrom Date: Tue, 3 May 2011 22:09:30 +0200 Subject: [PATCH 04/28] IPVS: Change of socket usage to enable name space exit. If the sync daemons run in a name space while it crashes or get killed, there is no way to stop them except for a reboot. When all patches are there, ip_vs_core will handle register_pernet_(), i.e. ip_vs_sync_init() and ip_vs_sync_cleanup() will be removed. Kernel threads should not increment the use count of a socket. By calling sk_change_net() after creating a socket this is avoided. sock_release cant be used intead sk_release_kernel() should be used. Thanks Eric W Biederman for your advices. Signed-off-by: Hans Schillstrom [horms@verge.net.au: minor edit to changelog] Signed-off-by: Simon Horman --- net/netfilter/ipvs/ip_vs_core.c | 2 +- net/netfilter/ipvs/ip_vs_sync.c | 58 +++++++++++++++++++++------------ 2 files changed, 38 insertions(+), 22 deletions(-) diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index 07accf6b2401..a0791dc05a27 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -1896,7 +1896,7 @@ static int __net_init __ip_vs_init(struct net *net) static void __net_exit __ip_vs_cleanup(struct net *net) { - IP_VS_DBG(10, "ipvs netns %d released\n", net_ipvs(net)->gen); + IP_VS_DBG(2, "ipvs netns %d released\n", net_ipvs(net)->gen); } static struct pernet_operations ipvs_core_ops = { diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index 3e7961e85e9c..0cce95310820 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -1303,13 +1303,18 @@ static struct socket *make_send_sock(struct net *net) struct socket *sock; int result; - /* First create a socket */ - result = __sock_create(net, PF_INET, SOCK_DGRAM, IPPROTO_UDP, &sock, 1); + /* First create a socket move it to right name space later */ + result = sock_create_kern(PF_INET, SOCK_DGRAM, IPPROTO_UDP, &sock); if (result < 0) { pr_err("Error during creation of socket; terminating\n"); return ERR_PTR(result); } - + /* + * Kernel sockets that are a part of a namespace, should not + * hold a reference to a namespace in order to allow to stop it. + * After sk_change_net should be released using sk_release_kernel. + */ + sk_change_net(sock->sk, net); result = set_mcast_if(sock->sk, ipvs->master_mcast_ifn); if (result < 0) { pr_err("Error setting outbound mcast interface\n"); @@ -1334,8 +1339,8 @@ static struct socket *make_send_sock(struct net *net) return sock; - error: - sock_release(sock); +error: + sk_release_kernel(sock->sk); return ERR_PTR(result); } @@ -1350,12 +1355,17 @@ static struct socket *make_receive_sock(struct net *net) int result; /* First create a socket */ - result = __sock_create(net, PF_INET, SOCK_DGRAM, IPPROTO_UDP, &sock, 1); + result = sock_create_kern(PF_INET, SOCK_DGRAM, IPPROTO_UDP, &sock); if (result < 0) { pr_err("Error during creation of socket; terminating\n"); return ERR_PTR(result); } - + /* + * Kernel sockets that are a part of a namespace, should not + * hold a reference to a namespace in order to allow to stop it. + * After sk_change_net should be released using sk_release_kernel. + */ + sk_change_net(sock->sk, net); /* it is equivalent to the REUSEADDR option in user-space */ sock->sk->sk_reuse = 1; @@ -1377,8 +1387,8 @@ static struct socket *make_receive_sock(struct net *net) return sock; - error: - sock_release(sock); +error: + sk_release_kernel(sock->sk); return ERR_PTR(result); } @@ -1473,7 +1483,7 @@ static int sync_thread_master(void *data) ip_vs_sync_buff_release(sb); /* release the sending multicast socket */ - sock_release(tinfo->sock); + sk_release_kernel(tinfo->sock->sk); kfree(tinfo); return 0; @@ -1513,7 +1523,7 @@ static int sync_thread_backup(void *data) } /* release the sending multicast socket */ - sock_release(tinfo->sock); + sk_release_kernel(tinfo->sock->sk); kfree(tinfo->buf); kfree(tinfo); @@ -1601,7 +1611,7 @@ outtinfo: outbuf: kfree(buf); outsocket: - sock_release(sock); + sk_release_kernel(sock->sk); out: return result; } @@ -1610,6 +1620,7 @@ out: int stop_sync_thread(struct net *net, int state) { struct netns_ipvs *ipvs = net_ipvs(net); + int retc = -EINVAL; IP_VS_DBG(7, "%s(): pid %d\n", __func__, task_pid_nr(current)); @@ -1629,7 +1640,7 @@ int stop_sync_thread(struct net *net, int state) spin_lock_bh(&ipvs->sync_lock); ipvs->sync_state &= ~IP_VS_STATE_MASTER; spin_unlock_bh(&ipvs->sync_lock); - kthread_stop(ipvs->master_thread); + retc = kthread_stop(ipvs->master_thread); ipvs->master_thread = NULL; } else if (state == IP_VS_STATE_BACKUP) { if (!ipvs->backup_thread) @@ -1639,16 +1650,14 @@ int stop_sync_thread(struct net *net, int state) task_pid_nr(ipvs->backup_thread)); ipvs->sync_state &= ~IP_VS_STATE_BACKUP; - kthread_stop(ipvs->backup_thread); + retc = kthread_stop(ipvs->backup_thread); ipvs->backup_thread = NULL; - } else { - return -EINVAL; } /* decrease the module use count */ ip_vs_use_count_dec(); - return 0; + return retc; } /* @@ -1670,8 +1679,15 @@ static int __net_init __ip_vs_sync_init(struct net *net) static void __ip_vs_sync_cleanup(struct net *net) { - stop_sync_thread(net, IP_VS_STATE_MASTER); - stop_sync_thread(net, IP_VS_STATE_BACKUP); + int retc; + + retc = stop_sync_thread(net, IP_VS_STATE_MASTER); + if (retc && retc != -ESRCH) + pr_err("Failed to stop Master Daemon\n"); + + retc = stop_sync_thread(net, IP_VS_STATE_BACKUP); + if (retc && retc != -ESRCH) + pr_err("Failed to stop Backup Daemon\n"); } static struct pernet_operations ipvs_sync_ops = { @@ -1682,10 +1698,10 @@ static struct pernet_operations ipvs_sync_ops = { int __init ip_vs_sync_init(void) { - return register_pernet_subsys(&ipvs_sync_ops); + return register_pernet_device(&ipvs_sync_ops); } void ip_vs_sync_cleanup(void) { - unregister_pernet_subsys(&ipvs_sync_ops); + unregister_pernet_device(&ipvs_sync_ops); } From 74973f6fbfcd1b084c3ccc75b783a6dacac94a10 Mon Sep 17 00:00:00 2001 From: Hans Schillstrom Date: Tue, 3 May 2011 22:09:31 +0200 Subject: [PATCH 05/28] IPVS: init and cleanup restructuring DESCRIPTION This patch tries to restore the initial init and cleanup sequences that was before namspace patch. Netns also requires action when net devices unregister which has never been implemented. I.e this patch also covers when a device moves into a network namespace, and has to be released. IMPLEMENTATION The number of calls to register_pernet_device have been reduced to one for the ip_vs.ko Schedulers still have their own calls. This patch adds a function __ip_vs_service_cleanup() and an enable flag for the netfilter hooks. The nf hooks will be enabled when the first service is loaded and never disabled again, except when a namespace exit starts. Signed-off-by: Hans Schillstrom Acked-by: Julian Anastasov [horms@verge.net.au: minor edit to changelog] Signed-off-by: Simon Horman --- include/net/ip_vs.h | 17 +++++ net/netfilter/ipvs/ip_vs_app.c | 15 +--- net/netfilter/ipvs/ip_vs_conn.c | 12 +--- net/netfilter/ipvs/ip_vs_core.c | 101 +++++++++++++++++++++++--- net/netfilter/ipvs/ip_vs_ctl.c | 120 ++++++++++++++++++++++++++----- net/netfilter/ipvs/ip_vs_est.c | 14 +--- net/netfilter/ipvs/ip_vs_proto.c | 11 +-- net/netfilter/ipvs/ip_vs_sync.c | 13 +--- 8 files changed, 223 insertions(+), 80 deletions(-) diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 4d1b71ae82ba..02f67021e079 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -791,6 +791,7 @@ struct ip_vs_app { /* IPVS in network namespace */ struct netns_ipvs { int gen; /* Generation */ + int enable; /* enable like nf_hooks do */ /* * Hash table: for real service lookups */ @@ -1089,6 +1090,22 @@ ip_vs_control_add(struct ip_vs_conn *cp, struct ip_vs_conn *ctl_cp) atomic_inc(&ctl_cp->n_control); } +/* + * IPVS netns init & cleanup functions + */ +extern int __ip_vs_estimator_init(struct net *net); +extern int __ip_vs_control_init(struct net *net); +extern int __ip_vs_protocol_init(struct net *net); +extern int __ip_vs_app_init(struct net *net); +extern int __ip_vs_conn_init(struct net *net); +extern int __ip_vs_sync_init(struct net *net); +extern void __ip_vs_conn_cleanup(struct net *net); +extern void __ip_vs_app_cleanup(struct net *net); +extern void __ip_vs_protocol_cleanup(struct net *net); +extern void __ip_vs_control_cleanup(struct net *net); +extern void __ip_vs_estimator_cleanup(struct net *net); +extern void __ip_vs_sync_cleanup(struct net *net); +extern void __ip_vs_service_cleanup(struct net *net); /* * IPVS application functions diff --git a/net/netfilter/ipvs/ip_vs_app.c b/net/netfilter/ipvs/ip_vs_app.c index 2dc6de13ac18..51f3af7c4743 100644 --- a/net/netfilter/ipvs/ip_vs_app.c +++ b/net/netfilter/ipvs/ip_vs_app.c @@ -576,7 +576,7 @@ static const struct file_operations ip_vs_app_fops = { }; #endif -static int __net_init __ip_vs_app_init(struct net *net) +int __net_init __ip_vs_app_init(struct net *net) { struct netns_ipvs *ipvs = net_ipvs(net); @@ -585,26 +585,17 @@ static int __net_init __ip_vs_app_init(struct net *net) return 0; } -static void __net_exit __ip_vs_app_cleanup(struct net *net) +void __net_exit __ip_vs_app_cleanup(struct net *net) { proc_net_remove(net, "ip_vs_app"); } -static struct pernet_operations ip_vs_app_ops = { - .init = __ip_vs_app_init, - .exit = __ip_vs_app_cleanup, -}; - int __init ip_vs_app_init(void) { - int rv; - - rv = register_pernet_subsys(&ip_vs_app_ops); - return rv; + return 0; } void ip_vs_app_cleanup(void) { - unregister_pernet_subsys(&ip_vs_app_ops); } diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c index f289306cbf12..5092505f27a7 100644 --- a/net/netfilter/ipvs/ip_vs_conn.c +++ b/net/netfilter/ipvs/ip_vs_conn.c @@ -1258,22 +1258,17 @@ int __net_init __ip_vs_conn_init(struct net *net) return 0; } -static void __net_exit __ip_vs_conn_cleanup(struct net *net) +void __net_exit __ip_vs_conn_cleanup(struct net *net) { /* flush all the connection entries first */ ip_vs_conn_flush(net); proc_net_remove(net, "ip_vs_conn"); proc_net_remove(net, "ip_vs_conn_sync"); } -static struct pernet_operations ipvs_conn_ops = { - .init = __ip_vs_conn_init, - .exit = __ip_vs_conn_cleanup, -}; int __init ip_vs_conn_init(void) { int idx; - int retc; /* Compute size and mask */ ip_vs_conn_tab_size = 1 << ip_vs_conn_tab_bits; @@ -1309,17 +1304,14 @@ int __init ip_vs_conn_init(void) rwlock_init(&__ip_vs_conntbl_lock_array[idx].l); } - retc = register_pernet_subsys(&ipvs_conn_ops); - /* calculate the random value for connection hash */ get_random_bytes(&ip_vs_conn_rnd, sizeof(ip_vs_conn_rnd)); - return retc; + return 0; } void ip_vs_conn_cleanup(void) { - unregister_pernet_subsys(&ipvs_conn_ops); /* Release the empty cache */ kmem_cache_destroy(ip_vs_conn_cachep); vfree(ip_vs_conn_tab); diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index a0791dc05a27..a74dae6c5dbc 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -1113,6 +1113,9 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af) return NF_ACCEPT; net = skb_net(skb); + if (!net_ipvs(net)->enable) + return NF_ACCEPT; + ip_vs_fill_iphdr(af, skb_network_header(skb), &iph); #ifdef CONFIG_IP_VS_IPV6 if (af == AF_INET6) { @@ -1343,6 +1346,7 @@ ip_vs_in_icmp(struct sk_buff *skb, int *related, unsigned int hooknum) return NF_ACCEPT; /* The packet looks wrong, ignore */ net = skb_net(skb); + pd = ip_vs_proto_data_get(net, cih->protocol); if (!pd) return NF_ACCEPT; @@ -1529,6 +1533,11 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af) IP_VS_DBG_ADDR(af, &iph.daddr), hooknum); return NF_ACCEPT; } + /* ipvs enabled in this netns ? */ + net = skb_net(skb); + if (!net_ipvs(net)->enable) + return NF_ACCEPT; + ip_vs_fill_iphdr(af, skb_network_header(skb), &iph); /* Bad... Do not break raw sockets */ @@ -1562,7 +1571,6 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af) ip_vs_fill_iphdr(af, skb_network_header(skb), &iph); } - net = skb_net(skb); /* Protocol supported? */ pd = ip_vs_proto_data_get(net, iph.protocol); if (unlikely(!pd)) @@ -1588,7 +1596,6 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af) } IP_VS_DBG_PKT(11, af, pp, skb, 0, "Incoming packet"); - net = skb_net(skb); ipvs = net_ipvs(net); /* Check the server status */ if (cp->dest && !(cp->dest->flags & IP_VS_DEST_F_AVAILABLE)) { @@ -1743,10 +1750,16 @@ ip_vs_forward_icmp(unsigned int hooknum, struct sk_buff *skb, int (*okfn)(struct sk_buff *)) { int r; + struct net *net; if (ip_hdr(skb)->protocol != IPPROTO_ICMP) return NF_ACCEPT; + /* ipvs enabled in this netns ? */ + net = skb_net(skb); + if (!net_ipvs(net)->enable) + return NF_ACCEPT; + return ip_vs_in_icmp(skb, &r, hooknum); } @@ -1757,10 +1770,16 @@ ip_vs_forward_icmp_v6(unsigned int hooknum, struct sk_buff *skb, int (*okfn)(struct sk_buff *)) { int r; + struct net *net; if (ipv6_hdr(skb)->nexthdr != IPPROTO_ICMPV6) return NF_ACCEPT; + /* ipvs enabled in this netns ? */ + net = skb_net(skb); + if (!net_ipvs(net)->enable) + return NF_ACCEPT; + return ip_vs_in_icmp_v6(skb, &r, hooknum); } #endif @@ -1884,21 +1903,72 @@ static int __net_init __ip_vs_init(struct net *net) pr_err("%s(): no memory.\n", __func__); return -ENOMEM; } + /* Hold the beast until a service is registerd */ + ipvs->enable = 0; ipvs->net = net; /* Counters used for creating unique names */ ipvs->gen = atomic_read(&ipvs_netns_cnt); atomic_inc(&ipvs_netns_cnt); net->ipvs = ipvs; + + if (__ip_vs_estimator_init(net) < 0) + goto estimator_fail; + + if (__ip_vs_control_init(net) < 0) + goto control_fail; + + if (__ip_vs_protocol_init(net) < 0) + goto protocol_fail; + + if (__ip_vs_app_init(net) < 0) + goto app_fail; + + if (__ip_vs_conn_init(net) < 0) + goto conn_fail; + + if (__ip_vs_sync_init(net) < 0) + goto sync_fail; + printk(KERN_INFO "IPVS: Creating netns size=%zu id=%d\n", sizeof(struct netns_ipvs), ipvs->gen); return 0; +/* + * Error handling + */ + +sync_fail: + __ip_vs_conn_cleanup(net); +conn_fail: + __ip_vs_app_cleanup(net); +app_fail: + __ip_vs_protocol_cleanup(net); +protocol_fail: + __ip_vs_control_cleanup(net); +control_fail: + __ip_vs_estimator_cleanup(net); +estimator_fail: + return -ENOMEM; } static void __net_exit __ip_vs_cleanup(struct net *net) { + __ip_vs_service_cleanup(net); /* ip_vs_flush() with locks */ + __ip_vs_conn_cleanup(net); + __ip_vs_app_cleanup(net); + __ip_vs_protocol_cleanup(net); + __ip_vs_control_cleanup(net); + __ip_vs_estimator_cleanup(net); IP_VS_DBG(2, "ipvs netns %d released\n", net_ipvs(net)->gen); } +static void __net_exit __ip_vs_dev_cleanup(struct net *net) +{ + EnterFunction(2); + net_ipvs(net)->enable = 0; /* Disable packet reception */ + __ip_vs_sync_cleanup(net); + LeaveFunction(2); +} + static struct pernet_operations ipvs_core_ops = { .init = __ip_vs_init, .exit = __ip_vs_cleanup, @@ -1906,6 +1976,10 @@ static struct pernet_operations ipvs_core_ops = { .size = sizeof(struct netns_ipvs), }; +static struct pernet_operations ipvs_core_dev_ops = { + .exit = __ip_vs_dev_cleanup, +}; + /* * Initialize IP Virtual Server */ @@ -1913,10 +1987,6 @@ static int __init ip_vs_init(void) { int ret; - ret = register_pernet_subsys(&ipvs_core_ops); /* Alloc ip_vs struct */ - if (ret < 0) - return ret; - ip_vs_estimator_init(); ret = ip_vs_control_init(); if (ret < 0) { @@ -1944,15 +2014,28 @@ static int __init ip_vs_init(void) goto cleanup_conn; } + ret = register_pernet_subsys(&ipvs_core_ops); /* Alloc ip_vs struct */ + if (ret < 0) + goto cleanup_sync; + + ret = register_pernet_device(&ipvs_core_dev_ops); + if (ret < 0) + goto cleanup_sub; + ret = nf_register_hooks(ip_vs_ops, ARRAY_SIZE(ip_vs_ops)); if (ret < 0) { pr_err("can't register hooks.\n"); - goto cleanup_sync; + goto cleanup_dev; } pr_info("ipvs loaded.\n"); + return ret; +cleanup_dev: + unregister_pernet_device(&ipvs_core_dev_ops); +cleanup_sub: + unregister_pernet_subsys(&ipvs_core_ops); cleanup_sync: ip_vs_sync_cleanup(); cleanup_conn: @@ -1964,20 +2047,20 @@ cleanup_sync: ip_vs_control_cleanup(); cleanup_estimator: ip_vs_estimator_cleanup(); - unregister_pernet_subsys(&ipvs_core_ops); /* free ip_vs struct */ return ret; } static void __exit ip_vs_cleanup(void) { nf_unregister_hooks(ip_vs_ops, ARRAY_SIZE(ip_vs_ops)); + unregister_pernet_device(&ipvs_core_dev_ops); + unregister_pernet_subsys(&ipvs_core_ops); /* free ip_vs struct */ ip_vs_sync_cleanup(); ip_vs_conn_cleanup(); ip_vs_app_cleanup(); ip_vs_protocol_cleanup(); ip_vs_control_cleanup(); ip_vs_estimator_cleanup(); - unregister_pernet_subsys(&ipvs_core_ops); /* free ip_vs struct */ pr_info("ipvs unloaded.\n"); } diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index ae47090bf45f..ea722810faf3 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -69,6 +69,11 @@ int ip_vs_get_debug_level(void) } #endif + +/* Protos */ +static void __ip_vs_del_service(struct ip_vs_service *svc); + + #ifdef CONFIG_IP_VS_IPV6 /* Taken from rt6_fill_node() in net/ipv6/route.c, is there a better way? */ static int __ip_vs_addr_is_local_v6(struct net *net, @@ -1214,6 +1219,8 @@ ip_vs_add_service(struct net *net, struct ip_vs_service_user_kern *u, write_unlock_bh(&__ip_vs_svc_lock); *svc_p = svc; + /* Now there is a service - full throttle */ + ipvs->enable = 1; return 0; @@ -1472,6 +1479,84 @@ static int ip_vs_flush(struct net *net) return 0; } +/* + * Delete service by {netns} in the service table. + * Called by __ip_vs_cleanup() + */ +void __ip_vs_service_cleanup(struct net *net) +{ + EnterFunction(2); + /* Check for "full" addressed entries */ + mutex_lock(&__ip_vs_mutex); + ip_vs_flush(net); + mutex_unlock(&__ip_vs_mutex); + LeaveFunction(2); +} +/* + * Release dst hold by dst_cache + */ +static inline void +__ip_vs_dev_reset(struct ip_vs_dest *dest, struct net_device *dev) +{ + spin_lock_bh(&dest->dst_lock); + if (dest->dst_cache && dest->dst_cache->dev == dev) { + IP_VS_DBG_BUF(3, "Reset dev:%s dest %s:%u ,dest->refcnt=%d\n", + dev->name, + IP_VS_DBG_ADDR(dest->af, &dest->addr), + ntohs(dest->port), + atomic_read(&dest->refcnt)); + ip_vs_dst_reset(dest); + } + spin_unlock_bh(&dest->dst_lock); + +} +/* + * Netdev event receiver + * Currently only NETDEV_UNREGISTER is handled, i.e. if we hold a reference to + * a device that is "unregister" it must be released. + */ +static int ip_vs_dst_event(struct notifier_block *this, unsigned long event, + void *ptr) +{ + struct net_device *dev = ptr; + struct net *net = dev_net(dev); + struct ip_vs_service *svc; + struct ip_vs_dest *dest; + unsigned int idx; + + if (event != NETDEV_UNREGISTER) + return NOTIFY_DONE; + IP_VS_DBG(3, "%s() dev=%s\n", __func__, dev->name); + EnterFunction(2); + mutex_lock(&__ip_vs_mutex); + for (idx = 0; idx < IP_VS_SVC_TAB_SIZE; idx++) { + list_for_each_entry(svc, &ip_vs_svc_table[idx], s_list) { + if (net_eq(svc->net, net)) { + list_for_each_entry(dest, &svc->destinations, + n_list) { + __ip_vs_dev_reset(dest, dev); + } + } + } + + list_for_each_entry(svc, &ip_vs_svc_fwm_table[idx], f_list) { + if (net_eq(svc->net, net)) { + list_for_each_entry(dest, &svc->destinations, + n_list) { + __ip_vs_dev_reset(dest, dev); + } + } + + } + } + + list_for_each_entry(dest, &net_ipvs(net)->dest_trash, n_list) { + __ip_vs_dev_reset(dest, dev); + } + mutex_unlock(&__ip_vs_mutex); + LeaveFunction(2); + return NOTIFY_DONE; +} /* * Zero counters in a service or all services @@ -3588,6 +3673,10 @@ void __net_init __ip_vs_control_cleanup_sysctl(struct net *net) { } #endif +static struct notifier_block ip_vs_dst_notifier = { + .notifier_call = ip_vs_dst_event, +}; + int __net_init __ip_vs_control_init(struct net *net) { int idx; @@ -3626,7 +3715,7 @@ err: return -ENOMEM; } -static void __net_exit __ip_vs_control_cleanup(struct net *net) +void __net_exit __ip_vs_control_cleanup(struct net *net) { struct netns_ipvs *ipvs = net_ipvs(net); @@ -3639,11 +3728,6 @@ static void __net_exit __ip_vs_control_cleanup(struct net *net) free_percpu(ipvs->tot_stats.cpustats); } -static struct pernet_operations ipvs_control_ops = { - .init = __ip_vs_control_init, - .exit = __ip_vs_control_cleanup, -}; - int __init ip_vs_control_init(void) { int idx; @@ -3657,33 +3741,32 @@ int __init ip_vs_control_init(void) INIT_LIST_HEAD(&ip_vs_svc_fwm_table[idx]); } - ret = register_pernet_subsys(&ipvs_control_ops); - if (ret) { - pr_err("cannot register namespace.\n"); - goto err; - } - smp_wmb(); /* Do we really need it now ? */ ret = nf_register_sockopt(&ip_vs_sockopts); if (ret) { pr_err("cannot register sockopt.\n"); - goto err_net; + goto err_sock; } ret = ip_vs_genl_register(); if (ret) { pr_err("cannot register Generic Netlink interface.\n"); - nf_unregister_sockopt(&ip_vs_sockopts); - goto err_net; + goto err_genl; } + ret = register_netdevice_notifier(&ip_vs_dst_notifier); + if (ret < 0) + goto err_notf; + LeaveFunction(2); return 0; -err_net: - unregister_pernet_subsys(&ipvs_control_ops); -err: +err_notf: + ip_vs_genl_unregister(); +err_genl: + nf_unregister_sockopt(&ip_vs_sockopts); +err_sock: return ret; } @@ -3691,7 +3774,6 @@ err: void ip_vs_control_cleanup(void) { EnterFunction(2); - unregister_pernet_subsys(&ipvs_control_ops); ip_vs_genl_unregister(); nf_unregister_sockopt(&ip_vs_sockopts); LeaveFunction(2); diff --git a/net/netfilter/ipvs/ip_vs_est.c b/net/netfilter/ipvs/ip_vs_est.c index 8c8766ca56ad..508cce98777c 100644 --- a/net/netfilter/ipvs/ip_vs_est.c +++ b/net/netfilter/ipvs/ip_vs_est.c @@ -192,7 +192,7 @@ void ip_vs_read_estimator(struct ip_vs_stats_user *dst, dst->outbps = (e->outbps + 0xF) >> 5; } -static int __net_init __ip_vs_estimator_init(struct net *net) +int __net_init __ip_vs_estimator_init(struct net *net) { struct netns_ipvs *ipvs = net_ipvs(net); @@ -203,24 +203,16 @@ static int __net_init __ip_vs_estimator_init(struct net *net) return 0; } -static void __net_exit __ip_vs_estimator_exit(struct net *net) +void __net_exit __ip_vs_estimator_cleanup(struct net *net) { del_timer_sync(&net_ipvs(net)->est_timer); } -static struct pernet_operations ip_vs_app_ops = { - .init = __ip_vs_estimator_init, - .exit = __ip_vs_estimator_exit, -}; int __init ip_vs_estimator_init(void) { - int rv; - - rv = register_pernet_subsys(&ip_vs_app_ops); - return rv; + return 0; } void ip_vs_estimator_cleanup(void) { - unregister_pernet_subsys(&ip_vs_app_ops); } diff --git a/net/netfilter/ipvs/ip_vs_proto.c b/net/netfilter/ipvs/ip_vs_proto.c index 17484a4416ef..eb86028536fc 100644 --- a/net/netfilter/ipvs/ip_vs_proto.c +++ b/net/netfilter/ipvs/ip_vs_proto.c @@ -316,7 +316,7 @@ ip_vs_tcpudp_debug_packet(int af, struct ip_vs_protocol *pp, /* * per network name-space init */ -static int __net_init __ip_vs_protocol_init(struct net *net) +int __net_init __ip_vs_protocol_init(struct net *net) { #ifdef CONFIG_IP_VS_PROTO_TCP register_ip_vs_proto_netns(net, &ip_vs_protocol_tcp); @@ -336,7 +336,7 @@ static int __net_init __ip_vs_protocol_init(struct net *net) return 0; } -static void __net_exit __ip_vs_protocol_cleanup(struct net *net) +void __net_exit __ip_vs_protocol_cleanup(struct net *net) { struct netns_ipvs *ipvs = net_ipvs(net); struct ip_vs_proto_data *pd; @@ -349,11 +349,6 @@ static void __net_exit __ip_vs_protocol_cleanup(struct net *net) } } -static struct pernet_operations ipvs_proto_ops = { - .init = __ip_vs_protocol_init, - .exit = __ip_vs_protocol_cleanup, -}; - int __init ip_vs_protocol_init(void) { char protocols[64]; @@ -382,7 +377,6 @@ int __init ip_vs_protocol_init(void) REGISTER_PROTOCOL(&ip_vs_protocol_esp); #endif pr_info("Registered protocols (%s)\n", &protocols[2]); - return register_pernet_subsys(&ipvs_proto_ops); return 0; } @@ -393,7 +387,6 @@ void ip_vs_protocol_cleanup(void) struct ip_vs_protocol *pp; int i; - unregister_pernet_subsys(&ipvs_proto_ops); /* unregister all the ipvs protocols */ for (i = 0; i < IP_VS_PROTO_TAB_SIZE; i++) { while ((pp = ip_vs_proto_table[i]) != NULL) diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index 0cce95310820..e292e5bddc70 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -1663,7 +1663,7 @@ int stop_sync_thread(struct net *net, int state) /* * Initialize data struct for each netns */ -static int __net_init __ip_vs_sync_init(struct net *net) +int __net_init __ip_vs_sync_init(struct net *net) { struct netns_ipvs *ipvs = net_ipvs(net); @@ -1677,7 +1677,7 @@ static int __net_init __ip_vs_sync_init(struct net *net) return 0; } -static void __ip_vs_sync_cleanup(struct net *net) +void __ip_vs_sync_cleanup(struct net *net) { int retc; @@ -1690,18 +1690,11 @@ static void __ip_vs_sync_cleanup(struct net *net) pr_err("Failed to stop Backup Daemon\n"); } -static struct pernet_operations ipvs_sync_ops = { - .init = __ip_vs_sync_init, - .exit = __ip_vs_sync_cleanup, -}; - - int __init ip_vs_sync_init(void) { - return register_pernet_device(&ipvs_sync_ops); + return 0; } void ip_vs_sync_cleanup(void) { - unregister_pernet_device(&ipvs_sync_ops); } From 5db1c07ced19b2eec3a149a3c624d88e02e246ae Mon Sep 17 00:00:00 2001 From: Luciano Coelho Date: Tue, 3 May 2011 21:40:08 +0300 Subject: [PATCH 06/28] mac80211: don't start the dynamic ps timer if not associated When we are disconnecting, we set PS off, but this happens before we send the deauth/disassoc request. When the deauth/disassoc frames are sent, we trigger the dynamic ps timer, which then times out and turns PS back on. Thus, PS remains on after disconnecting, causing problems when associating again. This can be fixed by preventing the timer to start when we're not associated anymore. Signed-off-by: Luciano Coelho Signed-off-by: John W. Linville --- net/mac80211/tx.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index ce4596ed1268..bd1224fd216a 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -237,6 +237,10 @@ ieee80211_tx_h_dynamic_ps(struct ieee80211_tx_data *tx) &local->dynamic_ps_disable_work); } + /* Don't restart the timer if we're not disassociated */ + if (!ifmgd->associated) + return TX_CONTINUE; + mod_timer(&local->dynamic_ps_timer, jiffies + msecs_to_jiffies(local->hw.conf.dynamic_ps_timeout)); From 99aa55b66e3553e6f7212ec1104e0fac06cc558e Mon Sep 17 00:00:00 2001 From: Mohammed Shafi Shajakhan Date: Fri, 6 May 2011 20:43:11 +0530 Subject: [PATCH 07/28] ath9k: Fix a warning due to a queued work during S3 state during suspend/S3 state drv_flush is called from mac80211 irrespective of interface count. In ath9k we queue a work in ath9k_flush which we expect to be cancelled in the drv_stop call back. during suspend process mac80211 calls drv_stop only when the interface count(local->count) is non-zero. unfortunately when the network manager is enabled, drv_flush is called while drv_stop is not called as local->count reaches '0'. So fix this by simply checking for the device presence in the drv_flush call back in the driver before queueing work or anything else. this patch fixes the following WARNING Call Trace: [] warn_slowpath_common+0x72/0xa0 [] ? ieee80211_can_queue_work+0x39/0x50 [mac80211] [] ? ieee80211_can_queue_work+0x39/0x50 [mac80211] [] warn_slowpath_fmt+0x2b/0x30 [] ieee80211_can_queue_work+0x39/0x50 [mac80211] [] ieee80211_queue_delayed_work+0x21/0x50 [mac80211] [] ath_tx_complete_poll_work+0xb2/0x100 [ath9k] [] run_workqueue+0x8e/0x150 [] ? ath_tx_complete_poll_work+0x0/0x100 [ath9k] [] worker_thread+0x84/0xe0 [] ? autoremove_wake_function+0x0/0x50 [] ? worker_thread+0x0/0xe0 [] kthread+0x74/0x80 [] ? kthread+0x0/0x80 [] kernel_thread_helper+0x7/0x10 ---[ end trace 2aff81010df9215b ]--- Signed-off-by: Rajkumar Manoharan Signed-off-by: Mohammed Shafi Shajakhan Signed-off-by: John W. Linville --- drivers/net/wireless/ath/ath9k/main.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c index 17d04ff8d678..1482fa650833 100644 --- a/drivers/net/wireless/ath/ath9k/main.c +++ b/drivers/net/wireless/ath/ath9k/main.c @@ -2141,6 +2141,8 @@ static void ath9k_set_coverage_class(struct ieee80211_hw *hw, u8 coverage_class) static void ath9k_flush(struct ieee80211_hw *hw, bool drop) { struct ath_softc *sc = hw->priv; + struct ath_hw *ah = sc->sc_ah; + struct ath_common *common = ath9k_hw_common(ah); int timeout = 200; /* ms */ int i, j; @@ -2149,6 +2151,12 @@ static void ath9k_flush(struct ieee80211_hw *hw, bool drop) cancel_delayed_work_sync(&sc->tx_complete_work); + if (sc->sc_flags & SC_OP_INVALID) { + ath_dbg(common, ATH_DBG_ANY, "Device not present\n"); + mutex_unlock(&sc->mutex); + return; + } + if (drop) timeout = 1; From eb85de3f84868ca85703a23617b4079ce79a801e Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Sat, 7 May 2011 17:46:21 +0200 Subject: [PATCH 08/28] iwlegacy: fix IBSS mode crashes We should not switch to non-IBSS channels when working in IBSS mode, otherwise there are microcode errors, and after some time system crashes. This bug is only observable when software scan is used in IBSS mode, so should be considered as regression after: commit 0263aa45293838b514b8af674a03faf040991a90 Author: Stanislaw Gruszka Date: Tue Mar 29 11:24:21 2011 +0200 iwl3945: disable hw scan by default However IBSS mode check, which this patch add again, was removed by commit b2f30e8bdd8ef5f3b5a7ef9146509585a15347d3 Author: Johannes Berg Date: Thu Jan 21 07:32:20 2010 -0800 iwlwifi: remove IBSS channel sanity check That commit claim that mac80211 will not use non-IBSS channel in IBSS mode, what definitely is not true. Bug probably should be fixed in mac80211, but that will require more work, so better to apply that patch temporally, and provide proper mac80211 fix latter. Resolves: https://bugzilla.kernel.org/show_bug.cgi?id=34452 Reported-and-tested-by: Mikko Rapeli Cc: stable@kernel.org # 2.6.38.5+ Signed-off-by: Stanislaw Gruszka Signed-off-by: John W. Linville --- drivers/net/wireless/iwlegacy/iwl-core.c | 7 +++++++ drivers/net/wireless/iwlegacy/iwl-dev.h | 6 ++++++ 2 files changed, 13 insertions(+) diff --git a/drivers/net/wireless/iwlegacy/iwl-core.c b/drivers/net/wireless/iwlegacy/iwl-core.c index 2b08efb3b658..dcbb2ef27f22 100644 --- a/drivers/net/wireless/iwlegacy/iwl-core.c +++ b/drivers/net/wireless/iwlegacy/iwl-core.c @@ -2155,6 +2155,13 @@ int iwl_legacy_mac_config(struct ieee80211_hw *hw, u32 changed) goto set_ch_out; } + if (priv->iw_mode == NL80211_IFTYPE_ADHOC && + !iwl_legacy_is_channel_ibss(ch_info)) { + IWL_DEBUG_MAC80211(priv, "leave - not IBSS channel\n"); + ret = -EINVAL; + goto set_ch_out; + } + spin_lock_irqsave(&priv->lock, flags); for_each_context(priv, ctx) { diff --git a/drivers/net/wireless/iwlegacy/iwl-dev.h b/drivers/net/wireless/iwlegacy/iwl-dev.h index 9ee849d669f3..f43ac1eb9014 100644 --- a/drivers/net/wireless/iwlegacy/iwl-dev.h +++ b/drivers/net/wireless/iwlegacy/iwl-dev.h @@ -1411,6 +1411,12 @@ iwl_legacy_is_channel_passive(const struct iwl_channel_info *ch) return (!(ch->flags & EEPROM_CHANNEL_ACTIVE)) ? 1 : 0; } +static inline int +iwl_legacy_is_channel_ibss(const struct iwl_channel_info *ch) +{ + return (ch->flags & EEPROM_CHANNEL_IBSS) ? 1 : 0; +} + static inline void __iwl_legacy_free_pages(struct iwl_priv *priv, struct page *page) { From 2ae1b8b35faba31a59b153cbad07f9c15de99740 Mon Sep 17 00:00:00 2001 From: Paul Fox Date: Mon, 9 May 2011 10:40:42 +0100 Subject: [PATCH 09/28] libertas: fix cmdpendingq locking We occasionally see list corruption using libertas. While we haven't been able to diagnose this precisely, we have spotted a possible cause: cmdpendingq is generally modified with driver_lock held. However, there are a couple of points where this is not the case. Fix up those operations to execute under the lock, it seems like the correct thing to do and will hopefully improve the situation. Signed-off-by: Paul Fox Signed-off-by: Daniel Drake Acked-by: Dan Williams Cc: stable@kernel.org Signed-off-by: John W. Linville --- drivers/net/wireless/libertas/cmd.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/libertas/cmd.c b/drivers/net/wireless/libertas/cmd.c index 7e8a658b7670..f3ac62431a30 100644 --- a/drivers/net/wireless/libertas/cmd.c +++ b/drivers/net/wireless/libertas/cmd.c @@ -1339,8 +1339,8 @@ int lbs_execute_next_command(struct lbs_private *priv) cpu_to_le16(PS_MODE_ACTION_EXIT_PS)) { lbs_deb_host( "EXEC_NEXT_CMD: ignore ENTER_PS cmd\n"); - list_del(&cmdnode->list); spin_lock_irqsave(&priv->driver_lock, flags); + list_del(&cmdnode->list); lbs_complete_command(priv, cmdnode, 0); spin_unlock_irqrestore(&priv->driver_lock, flags); @@ -1352,8 +1352,8 @@ int lbs_execute_next_command(struct lbs_private *priv) (priv->psstate == PS_STATE_PRE_SLEEP)) { lbs_deb_host( "EXEC_NEXT_CMD: ignore EXIT_PS cmd in sleep\n"); - list_del(&cmdnode->list); spin_lock_irqsave(&priv->driver_lock, flags); + list_del(&cmdnode->list); lbs_complete_command(priv, cmdnode, 0); spin_unlock_irqrestore(&priv->driver_lock, flags); priv->needtowakeup = 1; @@ -1366,7 +1366,9 @@ int lbs_execute_next_command(struct lbs_private *priv) "EXEC_NEXT_CMD: sending EXIT_PS\n"); } } + spin_lock_irqsave(&priv->driver_lock, flags); list_del(&cmdnode->list); + spin_unlock_irqrestore(&priv->driver_lock, flags); lbs_deb_host("EXEC_NEXT_CMD: sending command 0x%04x\n", le16_to_cpu(cmd->command)); lbs_submit_command(priv, cmdnode); From 21ccc7936dac5ca9b3e2838bbc112a60f34e18b3 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Tue, 10 May 2011 16:17:10 +0000 Subject: [PATCH 10/28] ehea: Fix memory hotplug oops The ehea driver oopses during memory hotplug if the ports are not up. A simple testcase: # ifconfig ethX down # echo offline > /sys/devices/system/memory/memory32/state Oops: Kernel access of bad area, sig: 11 [#1] last sysfs file: /sys/devices/system/memory/memory32/state REGS: c000000709393110 TRAP: 0300 Not tainted (2.6.39-rc2-01385-g7ef73bc-dirty) DAR: 0000000000000000, DSISR: 40000000 ... NIP [c000000000067c98] .__wake_up_common+0x48/0xf0 LR [c00000000006d034] .__wake_up+0x54/0x90 Call Trace: [c00000000006d034] .__wake_up+0x54/0x90 [d000000006bb6270] .ehea_rereg_mrs+0x140/0x730 [ehea] [d000000006bb69c4] .ehea_mem_notifier+0x164/0x170 [ehea] [c0000000006fc8a8] .notifier_call_chain+0x78/0xf0 [c0000000000b3d70] .__blocking_notifier_call_chain+0x70/0xb0 [c000000000458d78] .memory_notify+0x28/0x40 [c0000000001871d8] .remove_memory+0x208/0x6d0 [c000000000458264] .memory_section_action+0x94/0x140 [c0000000004583ec] .memory_block_change_state+0xdc/0x1d0 [c0000000004585cc] .store_mem_state+0xec/0x160 [c00000000044768c] .sysdev_store+0x3c/0x50 [c00000000020b48c] .sysfs_write_file+0xec/0x1f0 [c00000000018f86c] .vfs_write+0xec/0x1e0 [c00000000018fa88] .SyS_write+0x58/0xd0 To fix this, initialise the waitqueues during port probe instead of port open. Signed-off-by: Anton Blanchard Cc: stable@kernel.org Acked-by: Breno Leitao Signed-off-by: David S. Miller --- drivers/net/ehea/ehea_main.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/net/ehea/ehea_main.c b/drivers/net/ehea/ehea_main.c index 53c0f04b1b23..cf79cf759e13 100644 --- a/drivers/net/ehea/ehea_main.c +++ b/drivers/net/ehea/ehea_main.c @@ -2688,9 +2688,6 @@ static int ehea_open(struct net_device *dev) netif_start_queue(dev); } - init_waitqueue_head(&port->swqe_avail_wq); - init_waitqueue_head(&port->restart_wq); - mutex_unlock(&port->port_lock); return ret; @@ -3276,6 +3273,9 @@ struct ehea_port *ehea_setup_single_port(struct ehea_adapter *adapter, INIT_WORK(&port->reset_task, ehea_reset_port); + init_waitqueue_head(&port->swqe_avail_wq); + init_waitqueue_head(&port->restart_wq); + ret = register_netdev(dev); if (ret) { pr_err("register_netdev failed. ret=%d\n", ret); From 747df2258b1b9a2e25929ef496262c339c380009 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Wed, 11 May 2011 17:41:18 +0100 Subject: [PATCH 11/28] sfc: Always map MCDI shared memory as uncacheable We enabled write-combining for memory-mapped registers in commit 65f0b417dee94f779ce9b77102b7d73c93723b39, but inhibited it for the MCDI shared memory where this is not supported. However, write-combining mappings also allow read-reordering, which may also be a problem. I found that when an SFC9000-family controller is connected to an Intel 3000 chipset, and write-combining is enabled, the controller stops responding to PCIe read requests during driver initialisation while the driver is polling for completion of an MCDI command. This results in an NMI and system hang. Adding read memory barriers between all reads to the shared memory area appears to reduce but not eliminate the probability of this. We have not yet established whether this is a bug in our BIU or in the PCIe bridge. For now, work around by mapping the shared memory area separately. Signed-off-by: Ben Hutchings --- drivers/net/sfc/mcdi.c | 49 +++++++++++++++++++++++++---------------- drivers/net/sfc/nic.h | 2 ++ drivers/net/sfc/siena.c | 25 +++++++++++++++++---- 3 files changed, 53 insertions(+), 23 deletions(-) diff --git a/drivers/net/sfc/mcdi.c b/drivers/net/sfc/mcdi.c index d98479030ef2..3dd45ed61f0a 100644 --- a/drivers/net/sfc/mcdi.c +++ b/drivers/net/sfc/mcdi.c @@ -50,6 +50,20 @@ static inline struct efx_mcdi_iface *efx_mcdi(struct efx_nic *efx) return &nic_data->mcdi; } +static inline void +efx_mcdi_readd(struct efx_nic *efx, efx_dword_t *value, unsigned reg) +{ + struct siena_nic_data *nic_data = efx->nic_data; + value->u32[0] = (__force __le32)__raw_readl(nic_data->mcdi_smem + reg); +} + +static inline void +efx_mcdi_writed(struct efx_nic *efx, const efx_dword_t *value, unsigned reg) +{ + struct siena_nic_data *nic_data = efx->nic_data; + __raw_writel((__force u32)value->u32[0], nic_data->mcdi_smem + reg); +} + void efx_mcdi_init(struct efx_nic *efx) { struct efx_mcdi_iface *mcdi; @@ -70,8 +84,8 @@ static void efx_mcdi_copyin(struct efx_nic *efx, unsigned cmd, const u8 *inbuf, size_t inlen) { struct efx_mcdi_iface *mcdi = efx_mcdi(efx); - unsigned pdu = FR_CZ_MC_TREG_SMEM + MCDI_PDU(efx); - unsigned doorbell = FR_CZ_MC_TREG_SMEM + MCDI_DOORBELL(efx); + unsigned pdu = MCDI_PDU(efx); + unsigned doorbell = MCDI_DOORBELL(efx); unsigned int i; efx_dword_t hdr; u32 xflags, seqno; @@ -92,30 +106,28 @@ static void efx_mcdi_copyin(struct efx_nic *efx, unsigned cmd, MCDI_HEADER_SEQ, seqno, MCDI_HEADER_XFLAGS, xflags); - efx_writed(efx, &hdr, pdu); + efx_mcdi_writed(efx, &hdr, pdu); - for (i = 0; i < inlen; i += 4) { - _efx_writed(efx, *((__le32 *)(inbuf + i)), pdu + 4 + i); - /* use wmb() within loop to inhibit write combining */ - wmb(); - } + for (i = 0; i < inlen; i += 4) + efx_mcdi_writed(efx, (const efx_dword_t *)(inbuf + i), + pdu + 4 + i); /* ring the doorbell with a distinctive value */ - _efx_writed(efx, (__force __le32) 0x45789abc, doorbell); - wmb(); + EFX_POPULATE_DWORD_1(hdr, EFX_DWORD_0, 0x45789abc); + efx_mcdi_writed(efx, &hdr, doorbell); } static void efx_mcdi_copyout(struct efx_nic *efx, u8 *outbuf, size_t outlen) { struct efx_mcdi_iface *mcdi = efx_mcdi(efx); - unsigned int pdu = FR_CZ_MC_TREG_SMEM + MCDI_PDU(efx); + unsigned int pdu = MCDI_PDU(efx); int i; BUG_ON(atomic_read(&mcdi->state) == MCDI_STATE_QUIESCENT); BUG_ON(outlen & 3 || outlen >= 0x100); for (i = 0; i < outlen; i += 4) - *((__le32 *)(outbuf + i)) = _efx_readd(efx, pdu + 4 + i); + efx_mcdi_readd(efx, (efx_dword_t *)(outbuf + i), pdu + 4 + i); } static int efx_mcdi_poll(struct efx_nic *efx) @@ -123,7 +135,7 @@ static int efx_mcdi_poll(struct efx_nic *efx) struct efx_mcdi_iface *mcdi = efx_mcdi(efx); unsigned int time, finish; unsigned int respseq, respcmd, error; - unsigned int pdu = FR_CZ_MC_TREG_SMEM + MCDI_PDU(efx); + unsigned int pdu = MCDI_PDU(efx); unsigned int rc, spins; efx_dword_t reg; @@ -149,8 +161,7 @@ static int efx_mcdi_poll(struct efx_nic *efx) time = get_seconds(); - rmb(); - efx_readd(efx, ®, pdu); + efx_mcdi_readd(efx, ®, pdu); /* All 1's indicates that shared memory is in reset (and is * not a valid header). Wait for it to come out reset before @@ -177,7 +188,7 @@ static int efx_mcdi_poll(struct efx_nic *efx) respseq, mcdi->seqno); rc = EIO; } else if (error) { - efx_readd(efx, ®, pdu + 4); + efx_mcdi_readd(efx, ®, pdu + 4); switch (EFX_DWORD_FIELD(reg, EFX_DWORD_0)) { #define TRANSLATE_ERROR(name) \ case MC_CMD_ERR_ ## name: \ @@ -211,21 +222,21 @@ out: /* Test and clear MC-rebooted flag for this port/function */ int efx_mcdi_poll_reboot(struct efx_nic *efx) { - unsigned int addr = FR_CZ_MC_TREG_SMEM + MCDI_REBOOT_FLAG(efx); + unsigned int addr = MCDI_REBOOT_FLAG(efx); efx_dword_t reg; uint32_t value; if (efx_nic_rev(efx) < EFX_REV_SIENA_A0) return false; - efx_readd(efx, ®, addr); + efx_mcdi_readd(efx, ®, addr); value = EFX_DWORD_FIELD(reg, EFX_DWORD_0); if (value == 0) return 0; EFX_ZERO_DWORD(reg); - efx_writed(efx, ®, addr); + efx_mcdi_writed(efx, ®, addr); if (value == MC_STATUS_DWORD_ASSERT) return -EINTR; diff --git a/drivers/net/sfc/nic.h b/drivers/net/sfc/nic.h index a42db6e35be3..d91701abd331 100644 --- a/drivers/net/sfc/nic.h +++ b/drivers/net/sfc/nic.h @@ -143,10 +143,12 @@ static inline struct falcon_board *falcon_board(struct efx_nic *efx) /** * struct siena_nic_data - Siena NIC state * @mcdi: Management-Controller-to-Driver Interface + * @mcdi_smem: MCDI shared memory mapping. The mapping is always uncacheable. * @wol_filter_id: Wake-on-LAN packet filter id */ struct siena_nic_data { struct efx_mcdi_iface mcdi; + void __iomem *mcdi_smem; int wol_filter_id; }; diff --git a/drivers/net/sfc/siena.c b/drivers/net/sfc/siena.c index e4dd8986b1fe..837869b71db9 100644 --- a/drivers/net/sfc/siena.c +++ b/drivers/net/sfc/siena.c @@ -220,12 +220,26 @@ static int siena_probe_nic(struct efx_nic *efx) efx_reado(efx, ®, FR_AZ_CS_DEBUG); efx->net_dev->dev_id = EFX_OWORD_FIELD(reg, FRF_CZ_CS_PORT_NUM) - 1; + /* Initialise MCDI */ + nic_data->mcdi_smem = ioremap_nocache(efx->membase_phys + + FR_CZ_MC_TREG_SMEM, + FR_CZ_MC_TREG_SMEM_STEP * + FR_CZ_MC_TREG_SMEM_ROWS); + if (!nic_data->mcdi_smem) { + netif_err(efx, probe, efx->net_dev, + "could not map MCDI at %llx+%x\n", + (unsigned long long)efx->membase_phys + + FR_CZ_MC_TREG_SMEM, + FR_CZ_MC_TREG_SMEM_STEP * FR_CZ_MC_TREG_SMEM_ROWS); + rc = -ENOMEM; + goto fail1; + } efx_mcdi_init(efx); /* Recover from a failed assertion before probing */ rc = efx_mcdi_handle_assertion(efx); if (rc) - goto fail1; + goto fail2; /* Let the BMC know that the driver is now in charge of link and * filter settings. We must do this before we reset the NIC */ @@ -280,6 +294,7 @@ fail4: fail3: efx_mcdi_drv_attach(efx, false, NULL); fail2: + iounmap(nic_data->mcdi_smem); fail1: kfree(efx->nic_data); return rc; @@ -359,6 +374,8 @@ static int siena_init_nic(struct efx_nic *efx) static void siena_remove_nic(struct efx_nic *efx) { + struct siena_nic_data *nic_data = efx->nic_data; + efx_nic_free_buffer(efx, &efx->irq_status); siena_reset_hw(efx, RESET_TYPE_ALL); @@ -368,7 +385,8 @@ static void siena_remove_nic(struct efx_nic *efx) efx_mcdi_drv_attach(efx, false, NULL); /* Tear down the private nic state */ - kfree(efx->nic_data); + iounmap(nic_data->mcdi_smem); + kfree(nic_data); efx->nic_data = NULL; } @@ -606,8 +624,7 @@ struct efx_nic_type siena_a0_nic_type = { .default_mac_ops = &efx_mcdi_mac_operations, .revision = EFX_REV_SIENA_A0, - .mem_map_size = (FR_CZ_MC_TREG_SMEM + - FR_CZ_MC_TREG_SMEM_STEP * FR_CZ_MC_TREG_SMEM_ROWS), + .mem_map_size = FR_CZ_MC_TREG_SMEM, /* MC_TREG_SMEM mapped separately */ .txd_ptr_tbl_base = FR_BZ_TX_DESC_PTR_TBL, .rxd_ptr_tbl_base = FR_BZ_RX_DESC_PTR_TBL, .buf_tbl_base = FR_BZ_BUF_FULL_TBL, From cf7e032fc87d59c475df26c4d40bf45d401b2adb Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 12 May 2011 09:11:38 +0000 Subject: [PATCH 12/28] zorro8390: Fix regression caused during net_device_ops conversion Changeset b6114794a1c394534659f4a17420e48cf23aa922 ("zorro8390: convert to net_device_ops") broke zorro8390 by adding 8390.o to the link. That meant that lib8390.c was included twice, once in zorro8390.c and once in 8390.c, subject to different macros. This patch reverts that by avoiding the wrappers in 8390.c. Fix based on commits 217cbfa856dc1cbc2890781626c4032d9e3ec59f ("mac8390: fix regression caused during net_device_ops conversion") and 4e0168fa4842e27795a75b205a510f25b62181d9 ("mac8390: fix build with NET_POLL_CONTROLLER"). Reported-by: Christian T. Steigies Suggested-by: Finn Thain Signed-off-by: Geert Uytterhoeven Tested-by: Christian T. Steigies Cc: stable@kernel.org Signed-off-by: David S. Miller --- drivers/net/Makefile | 2 +- drivers/net/zorro8390.c | 12 ++++++------ 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/net/Makefile b/drivers/net/Makefile index 01b604ad155e..c64675f6c977 100644 --- a/drivers/net/Makefile +++ b/drivers/net/Makefile @@ -219,7 +219,7 @@ obj-$(CONFIG_SC92031) += sc92031.o obj-$(CONFIG_LP486E) += lp486e.o obj-$(CONFIG_ETH16I) += eth16i.o -obj-$(CONFIG_ZORRO8390) += zorro8390.o 8390.o +obj-$(CONFIG_ZORRO8390) += zorro8390.o obj-$(CONFIG_HPLANCE) += hplance.o 7990.o obj-$(CONFIG_MVME147_NET) += mvme147.o 7990.o obj-$(CONFIG_EQUALIZER) += eql.o diff --git a/drivers/net/zorro8390.c b/drivers/net/zorro8390.c index b78a38d9172a..8c7c522a056a 100644 --- a/drivers/net/zorro8390.c +++ b/drivers/net/zorro8390.c @@ -126,7 +126,7 @@ static int __devinit zorro8390_init_one(struct zorro_dev *z, board = z->resource.start; ioaddr = board+cards[i].offset; - dev = alloc_ei_netdev(); + dev = ____alloc_ei_netdev(0); if (!dev) return -ENOMEM; if (!request_mem_region(ioaddr, NE_IO_EXTENT*2, DRV_NAME)) { @@ -146,15 +146,15 @@ static int __devinit zorro8390_init_one(struct zorro_dev *z, static const struct net_device_ops zorro8390_netdev_ops = { .ndo_open = zorro8390_open, .ndo_stop = zorro8390_close, - .ndo_start_xmit = ei_start_xmit, - .ndo_tx_timeout = ei_tx_timeout, - .ndo_get_stats = ei_get_stats, - .ndo_set_multicast_list = ei_set_multicast_list, + .ndo_start_xmit = __ei_start_xmit, + .ndo_tx_timeout = __ei_tx_timeout, + .ndo_get_stats = __ei_get_stats, + .ndo_set_multicast_list = __ei_set_multicast_list, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, .ndo_change_mtu = eth_change_mtu, #ifdef CONFIG_NET_POLL_CONTROLLER - .ndo_poll_controller = ei_poll, + .ndo_poll_controller = __ei_poll, #endif }; From 0b25e0157dfa236a0629c16c8ad6f222f633f682 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 12 May 2011 09:11:39 +0000 Subject: [PATCH 13/28] hydra: Fix regression caused during net_device_ops conversion Changeset 5618f0d1193d6b051da9b59b0e32ad24397f06a4 ("hydra: convert to net_device_ops") broke hydra by adding 8390.o to the link. That meant that lib8390.c was included twice, once in hydra.c and once in 8390.c, subject to different macros. This patch reverts that by avoiding the wrappers in 8390.c. Fix based on commits 217cbfa856dc1cbc2890781626c4032d9e3ec59f ("mac8390: fix regression caused during net_device_ops conversion") and 4e0168fa4842e27795a75b205a510f25b62181d9 ("mac8390: fix build with NET_POLL_CONTROLLER"). Signed-off-by: Geert Uytterhoeven Cc: stable@kernel.org Signed-off-by: David S. Miller --- drivers/net/Makefile | 2 +- drivers/net/hydra.c | 14 +++++++------- 2 files changed, 8 insertions(+), 8 deletions(-) diff --git a/drivers/net/Makefile b/drivers/net/Makefile index c64675f6c977..4d2f09460d60 100644 --- a/drivers/net/Makefile +++ b/drivers/net/Makefile @@ -231,7 +231,7 @@ obj-$(CONFIG_SGI_IOC3_ETH) += ioc3-eth.o obj-$(CONFIG_DECLANCE) += declance.o obj-$(CONFIG_ATARILANCE) += atarilance.o obj-$(CONFIG_A2065) += a2065.o -obj-$(CONFIG_HYDRA) += hydra.o 8390.o +obj-$(CONFIG_HYDRA) += hydra.o obj-$(CONFIG_ARIADNE) += ariadne.o obj-$(CONFIG_CS89x0) += cs89x0.o obj-$(CONFIG_MACSONIC) += macsonic.o diff --git a/drivers/net/hydra.c b/drivers/net/hydra.c index c5ef62ceb840..1cd481c04202 100644 --- a/drivers/net/hydra.c +++ b/drivers/net/hydra.c @@ -98,15 +98,15 @@ static const struct net_device_ops hydra_netdev_ops = { .ndo_open = hydra_open, .ndo_stop = hydra_close, - .ndo_start_xmit = ei_start_xmit, - .ndo_tx_timeout = ei_tx_timeout, - .ndo_get_stats = ei_get_stats, - .ndo_set_multicast_list = ei_set_multicast_list, + .ndo_start_xmit = __ei_start_xmit, + .ndo_tx_timeout = __ei_tx_timeout, + .ndo_get_stats = __ei_get_stats, + .ndo_set_multicast_list = __ei_set_multicast_list, .ndo_validate_addr = eth_validate_addr, - .ndo_set_mac_address = eth_mac_addr, + .ndo_set_mac_address = eth_mac_addr, .ndo_change_mtu = eth_change_mtu, #ifdef CONFIG_NET_POLL_CONTROLLER - .ndo_poll_controller = ei_poll, + .ndo_poll_controller = __ei_poll, #endif }; @@ -125,7 +125,7 @@ static int __devinit hydra_init(struct zorro_dev *z) 0x10, 0x12, 0x14, 0x16, 0x18, 0x1a, 0x1c, 0x1e, }; - dev = alloc_ei_netdev(); + dev = ____alloc_ei_netdev(0); if (!dev) return -ENOMEM; From 2592a7354092afd304a8c067319b15ab1e441e35 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 12 May 2011 09:11:40 +0000 Subject: [PATCH 14/28] ne-h8300: Fix regression caused during net_device_ops conversion Changeset dcd39c90290297f6e6ed8a04bb20da7ac2b043c5 ("ne-h8300: convert to net_device_ops") broke ne-h8300 by adding 8390.o to the link. That meant that lib8390.c was included twice, once in ne-h8300.c and once in 8390.c, subject to different macros. This patch reverts that by avoiding the wrappers in 8390.c. Fix based on commits 217cbfa856dc1cbc2890781626c4032d9e3ec59f ("mac8390: fix regression caused during net_device_ops conversion") and 4e0168fa4842e27795a75b205a510f25b62181d9 ("mac8390: fix build with NET_POLL_CONTROLLER"). Signed-off-by: Geert Uytterhoeven Cc: stable@kernel.org Signed-off-by: David S. Miller --- drivers/net/Makefile | 2 +- drivers/net/ne-h8300.c | 16 ++++++++-------- 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/drivers/net/Makefile b/drivers/net/Makefile index 4d2f09460d60..e5a7375685ad 100644 --- a/drivers/net/Makefile +++ b/drivers/net/Makefile @@ -144,7 +144,7 @@ obj-$(CONFIG_NE3210) += ne3210.o 8390.o obj-$(CONFIG_SB1250_MAC) += sb1250-mac.o obj-$(CONFIG_B44) += b44.o obj-$(CONFIG_FORCEDETH) += forcedeth.o -obj-$(CONFIG_NE_H8300) += ne-h8300.o 8390.o +obj-$(CONFIG_NE_H8300) += ne-h8300.o obj-$(CONFIG_AX88796) += ax88796.o obj-$(CONFIG_BCM63XX_ENET) += bcm63xx_enet.o obj-$(CONFIG_FTMAC100) += ftmac100.o diff --git a/drivers/net/ne-h8300.c b/drivers/net/ne-h8300.c index 30be8c634ebd..7298a34bc795 100644 --- a/drivers/net/ne-h8300.c +++ b/drivers/net/ne-h8300.c @@ -167,7 +167,7 @@ static void cleanup_card(struct net_device *dev) #ifndef MODULE struct net_device * __init ne_probe(int unit) { - struct net_device *dev = alloc_ei_netdev(); + struct net_device *dev = ____alloc_ei_netdev(0); int err; if (!dev) @@ -197,15 +197,15 @@ static const struct net_device_ops ne_netdev_ops = { .ndo_open = ne_open, .ndo_stop = ne_close, - .ndo_start_xmit = ei_start_xmit, - .ndo_tx_timeout = ei_tx_timeout, - .ndo_get_stats = ei_get_stats, - .ndo_set_multicast_list = ei_set_multicast_list, + .ndo_start_xmit = __ei_start_xmit, + .ndo_tx_timeout = __ei_tx_timeout, + .ndo_get_stats = __ei_get_stats, + .ndo_set_multicast_list = __ei_set_multicast_list, .ndo_validate_addr = eth_validate_addr, - .ndo_set_mac_address = eth_mac_addr, + .ndo_set_mac_address = eth_mac_addr, .ndo_change_mtu = eth_change_mtu, #ifdef CONFIG_NET_POLL_CONTROLLER - .ndo_poll_controller = ei_poll, + .ndo_poll_controller = __ei_poll, #endif }; @@ -637,7 +637,7 @@ int init_module(void) int err; for (this_dev = 0; this_dev < MAX_NE_CARDS; this_dev++) { - struct net_device *dev = alloc_ei_netdev(); + struct net_device *dev = ____alloc_ei_netdev(0); if (!dev) break; if (io[this_dev]) { From ca06707022d6ba4744198a8ebbe4994786b0c613 Mon Sep 17 00:00:00 2001 From: "Steinar H. Gunderson" Date: Fri, 6 May 2011 23:44:46 +0000 Subject: [PATCH 15/28] ipv6: restore correct ECN handling on TCP xmit Since commit e9df2e8fd8fbc9 (Use appropriate sock tclass setting for routing lookup) we lost ability to properly add ECN codemarks to ipv6 TCP frames. It seems like TCP_ECN_send() calls INET_ECN_xmit(), which only sets the ECN bit in the IPv4 ToS field (inet_sk(sk)->tos), but after the patch, what's checked is inet6_sk(sk)->tclass, which is a completely different field. Close bug https://bugzilla.kernel.org/show_bug.cgi?id=34322 [Eric Dumazet] : added the INET_ECN_dontxmit() fix and replace macros by inline functions for clarity. Signed-off-by: Steinar H. Gunderson Signed-off-by: Eric Dumazet Cc: YOSHIFUJI Hideaki Cc: Andrew Morton Signed-off-by: David S. Miller --- include/net/inet_ecn.h | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/include/net/inet_ecn.h b/include/net/inet_ecn.h index 88bdd010d65d..2fa8d1341a0a 100644 --- a/include/net/inet_ecn.h +++ b/include/net/inet_ecn.h @@ -38,9 +38,19 @@ static inline __u8 INET_ECN_encapsulate(__u8 outer, __u8 inner) return outer; } -#define INET_ECN_xmit(sk) do { inet_sk(sk)->tos |= INET_ECN_ECT_0; } while (0) -#define INET_ECN_dontxmit(sk) \ - do { inet_sk(sk)->tos &= ~INET_ECN_MASK; } while (0) +static inline void INET_ECN_xmit(struct sock *sk) +{ + inet_sk(sk)->tos |= INET_ECN_ECT_0; + if (inet6_sk(sk) != NULL) + inet6_sk(sk)->tclass |= INET_ECN_ECT_0; +} + +static inline void INET_ECN_dontxmit(struct sock *sk) +{ + inet_sk(sk)->tos &= ~INET_ECN_MASK; + if (inet6_sk(sk) != NULL) + inet6_sk(sk)->tclass &= ~INET_ECN_MASK; +} #define IP6_ECN_flow_init(label) do { \ (label) &= ~htonl(INET_ECN_MASK << 20); \ From a10e14667635dde504ed9e7ee851494c2cf2ae8e Mon Sep 17 00:00:00 2001 From: Vitalii Demianets Date: Thu, 12 May 2011 23:04:29 +0000 Subject: [PATCH 16/28] bonding,llc: Fix structure sizeof incompatibility for some PDUs With some combinations of arch/compiler (e.g. arm-linux-gcc) the sizeof operator on structure returns value greater than expected. In cases when the structure is used for mapping PDU fields it may lead to unexpected results (such as holes and alignment problems in skb data). __packed prevents this undesired behavior. Signed-off-by: Vitalii Demianets Signed-off-by: David S. Miller --- drivers/net/bonding/bond_3ad.h | 10 +++++----- include/net/llc_pdu.h | 8 ++++---- 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/drivers/net/bonding/bond_3ad.h b/drivers/net/bonding/bond_3ad.h index b28baff70864..01b8a6af275b 100644 --- a/drivers/net/bonding/bond_3ad.h +++ b/drivers/net/bonding/bond_3ad.h @@ -39,7 +39,7 @@ typedef struct mac_addr { u8 mac_addr_value[ETH_ALEN]; -} mac_addr_t; +} __packed mac_addr_t; enum { BOND_AD_STABLE = 0, @@ -134,12 +134,12 @@ typedef struct lacpdu { u8 tlv_type_terminator; // = terminator u8 terminator_length; // = 0 u8 reserved_50[50]; // = 0 -} lacpdu_t; +} __packed lacpdu_t; typedef struct lacpdu_header { struct ethhdr hdr; struct lacpdu lacpdu; -} lacpdu_header_t; +} __packed lacpdu_header_t; // Marker Protocol Data Unit(PDU) structure(43.5.3.2 in the 802.3ad standard) typedef struct bond_marker { @@ -155,12 +155,12 @@ typedef struct bond_marker { u8 tlv_type_terminator; // = 0x00 u8 terminator_length; // = 0x00 u8 reserved_90[90]; // = 0 -} bond_marker_t; +} __packed bond_marker_t; typedef struct bond_marker_header { struct ethhdr hdr; struct bond_marker marker; -} bond_marker_header_t; +} __packed bond_marker_header_t; #pragma pack() diff --git a/include/net/llc_pdu.h b/include/net/llc_pdu.h index 75b8e2968c9b..f57e7d46a453 100644 --- a/include/net/llc_pdu.h +++ b/include/net/llc_pdu.h @@ -199,7 +199,7 @@ struct llc_pdu_sn { u8 ssap; u8 ctrl_1; u8 ctrl_2; -}; +} __packed; static inline struct llc_pdu_sn *llc_pdu_sn_hdr(struct sk_buff *skb) { @@ -211,7 +211,7 @@ struct llc_pdu_un { u8 dsap; u8 ssap; u8 ctrl_1; -}; +} __packed; static inline struct llc_pdu_un *llc_pdu_un_hdr(struct sk_buff *skb) { @@ -359,7 +359,7 @@ struct llc_xid_info { u8 fmt_id; /* always 0x81 for LLC */ u8 type; /* different if NULL/non-NULL LSAP */ u8 rw; /* sender receive window */ -}; +} __packed; /** * llc_pdu_init_as_xid_cmd - sets bytes 3, 4 & 5 of LLC header as XID @@ -415,7 +415,7 @@ struct llc_frmr_info { u8 curr_ssv; /* current send state variable val */ u8 curr_rsv; /* current receive state variable */ u8 ind_bits; /* indicator bits set with macro */ -}; +} __packed; extern void llc_pdu_set_cmd_rsp(struct sk_buff *skb, u8 type); extern void llc_pdu_set_pf_bit(struct sk_buff *skb, u8 bit_value); From cb68552858c64db302771469b1202ea09e696329 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Fri, 13 May 2011 16:03:24 -0400 Subject: [PATCH 17/28] bridge: fix forwarding of IPv6 The commit 6b1e960fdbd75dcd9bcc3ba5ff8898ff1ad30b6e bridge: Reset IPCB when entering IP stack on NF_FORWARD broke forwarding of IPV6 packets in bridge because it would call bp_parse_ip_options with an IPV6 packet. Reported-by: Noah Meyerhans Signed-off-by: Stephen Hemminger Reviewed-by: Eric Dumazet Signed-off-by: David S. Miller --- net/bridge/br_netfilter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/bridge/br_netfilter.c b/net/bridge/br_netfilter.c index f3bc322c5891..74ef4d4846a4 100644 --- a/net/bridge/br_netfilter.c +++ b/net/bridge/br_netfilter.c @@ -737,7 +737,7 @@ static unsigned int br_nf_forward_ip(unsigned int hook, struct sk_buff *skb, nf_bridge->mask |= BRNF_PKT_TYPE; } - if (br_parse_ip_options(skb)) + if (pf == PF_INET && br_parse_ip_options(skb)) return NF_DROP; /* The physdev module checks on this */ From d8083deb4f1aa0977980dfb834fcc336ef38318f Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Fri, 13 May 2011 16:03:24 -0400 Subject: [PATCH 18/28] bridge: fix forwarding of IPv6 The commit 6b1e960fdbd75dcd9bcc3ba5ff8898ff1ad30b6e bridge: Reset IPCB when entering IP stack on NF_FORWARD broke forwarding of IPV6 packets in bridge because it would call bp_parse_ip_options with an IPV6 packet. Reported-by: Noah Meyerhans Signed-off-by: Stephen Hemminger Reviewed-by: Eric Dumazet Signed-off-by: David S. Miller Signed-off-by: Pablo Neira Ayuso --- net/bridge/br_netfilter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/bridge/br_netfilter.c b/net/bridge/br_netfilter.c index f3bc322c5891..74ef4d4846a4 100644 --- a/net/bridge/br_netfilter.c +++ b/net/bridge/br_netfilter.c @@ -737,7 +737,7 @@ static unsigned int br_nf_forward_ip(unsigned int hook, struct sk_buff *skb, nf_bridge->mask |= BRNF_PKT_TYPE; } - if (br_parse_ip_options(skb)) + if (pf == PF_INET && br_parse_ip_options(skb)) return NF_DROP; /* The physdev module checks on this */ From 0f08190fe8af3cdb6ba19690eb0fa253ecef4bde Mon Sep 17 00:00:00 2001 From: Hans Schillstrom Date: Sun, 15 May 2011 17:20:29 +0200 Subject: [PATCH 19/28] IPVS: fix netns if reading ip_vs_* procfs entries Without this patch every access to ip_vs in procfs will increase the netns count i.e. an unbalanced get_net()/put_net(). (ipvsadm commands also use procfs.) The result is you can't exit a netns if reading ip_vs_* procfs entries. Signed-off-by: Hans Schillstrom Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_app.c | 2 +- net/netfilter/ipvs/ip_vs_conn.c | 4 ++-- net/netfilter/ipvs/ip_vs_ctl.c | 6 +++--- 3 files changed, 6 insertions(+), 6 deletions(-) diff --git a/net/netfilter/ipvs/ip_vs_app.c b/net/netfilter/ipvs/ip_vs_app.c index 51f3af7c4743..059af3120be7 100644 --- a/net/netfilter/ipvs/ip_vs_app.c +++ b/net/netfilter/ipvs/ip_vs_app.c @@ -572,7 +572,7 @@ static const struct file_operations ip_vs_app_fops = { .open = ip_vs_app_open, .read = seq_read, .llseek = seq_lseek, - .release = seq_release, + .release = seq_release_net, }; #endif diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c index d3fd91bbba49..bf28ac2fc99b 100644 --- a/net/netfilter/ipvs/ip_vs_conn.c +++ b/net/netfilter/ipvs/ip_vs_conn.c @@ -1046,7 +1046,7 @@ static const struct file_operations ip_vs_conn_fops = { .open = ip_vs_conn_open, .read = seq_read, .llseek = seq_lseek, - .release = seq_release, + .release = seq_release_net, }; static const char *ip_vs_origin_name(unsigned flags) @@ -1114,7 +1114,7 @@ static const struct file_operations ip_vs_conn_sync_fops = { .open = ip_vs_conn_sync_open, .read = seq_read, .llseek = seq_lseek, - .release = seq_release, + .release = seq_release_net, }; #endif diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index ea722810faf3..37890f228b19 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -2066,7 +2066,7 @@ static const struct file_operations ip_vs_info_fops = { .open = ip_vs_info_open, .read = seq_read, .llseek = seq_lseek, - .release = seq_release_private, + .release = seq_release_net, }; #endif @@ -2109,7 +2109,7 @@ static const struct file_operations ip_vs_stats_fops = { .open = ip_vs_stats_seq_open, .read = seq_read, .llseek = seq_lseek, - .release = single_release, + .release = single_release_net, }; static int ip_vs_stats_percpu_show(struct seq_file *seq, void *v) @@ -2178,7 +2178,7 @@ static const struct file_operations ip_vs_stats_percpu_fops = { .open = ip_vs_stats_percpu_seq_open, .read = seq_read, .llseek = seq_lseek, - .release = single_release, + .release = single_release_net, }; #endif From 274ea0e2a4cdf18110e5931b8ecbfef6353e5293 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Mon, 16 May 2011 14:42:26 +0200 Subject: [PATCH 20/28] netfilter: nf_ct_sip: validate Content-Length in TCP SIP messages Verify that the message length of a single SIP message, which is calculated based on the Content-Length field contained in the SIP message, does not exceed the packet boundaries. Signed-off-by: Patrick McHardy --- net/netfilter/nf_conntrack_sip.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c index bcf47eb518ef..1f81abde131f 100644 --- a/net/netfilter/nf_conntrack_sip.c +++ b/net/netfilter/nf_conntrack_sip.c @@ -1461,6 +1461,8 @@ static int sip_help_tcp(struct sk_buff *skb, unsigned int protoff, end += strlen("\r\n\r\n") + clen; msglen = origlen = end - dptr; + if (msglen > datalen) + return NF_DROP; ret = process_sip_msg(skb, ct, dataoff, &dptr, &msglen); if (ret != NF_ACCEPT) From e6e4d9ed11fb1fab8b3256a3dc14d71b5e984ac4 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Mon, 16 May 2011 14:45:39 +0200 Subject: [PATCH 21/28] netfilter: nf_ct_sip: fix SDP parsing in TCP SIP messages for some Cisco phones Some Cisco phones do not place the Content-Length field at the end of the SIP message. This is valid, due to a misunderstanding of the specification the parser expects the SDP body to start directly after the Content-Length field. Fix the parser to scan for \r\n\r\n to locate the beginning of the SDP body. Reported-by: Teresa Kang Signed-off-by: Patrick McHardy --- net/netfilter/nf_conntrack_sip.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c index 1f81abde131f..c05c0dc33499 100644 --- a/net/netfilter/nf_conntrack_sip.c +++ b/net/netfilter/nf_conntrack_sip.c @@ -1419,6 +1419,7 @@ static int sip_help_tcp(struct sk_buff *skb, unsigned int protoff, const char *dptr, *end; s16 diff, tdiff = 0; int ret = NF_ACCEPT; + bool term; typeof(nf_nat_sip_seq_adjust_hook) nf_nat_sip_seq_adjust; if (ctinfo != IP_CT_ESTABLISHED && @@ -1453,10 +1454,15 @@ static int sip_help_tcp(struct sk_buff *skb, unsigned int protoff, if (dptr + matchoff == end) break; - if (end + strlen("\r\n\r\n") > dptr + datalen) - break; - if (end[0] != '\r' || end[1] != '\n' || - end[2] != '\r' || end[3] != '\n') + term = false; + for (; end + strlen("\r\n\r\n") <= dptr + datalen; end++) { + if (end[0] == '\r' && end[1] == '\n' && + end[2] == '\r' && end[3] == '\n') { + term = true; + break; + } + } + if (!term) break; end += strlen("\r\n\r\n") + clen; From 867955f5682f7157fdafe8670804b9f8ea077bc7 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Mon, 16 May 2011 06:13:49 +0000 Subject: [PATCH 22/28] sfc: Fix oops in register dump after mapping change Commit 747df2258b1b9a2e25929ef496262c339c380009 ('sfc: Always map MCDI shared memory as uncacheable') introduced a separate mapping for the MCDI shared memory (MC_TREG_SMEM). This means we can no longer easily include it in the register dump. Since it is not particularly useful in debugging, substitute a recognisable dummy value. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller --- drivers/net/sfc/nic.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/net/sfc/nic.c b/drivers/net/sfc/nic.c index 10f1cb79c147..9b29a8d7c449 100644 --- a/drivers/net/sfc/nic.c +++ b/drivers/net/sfc/nic.c @@ -1937,6 +1937,13 @@ void efx_nic_get_regs(struct efx_nic *efx, void *buf) size = min_t(size_t, table->step, 16); + if (table->offset >= efx->type->mem_map_size) { + /* No longer mapped; return dummy data */ + memcpy(buf, "\xde\xc0\xad\xde", 4); + buf += table->rows * size; + continue; + } + for (i = 0; i < table->rows; i++) { switch (table->step) { case 4: /* 32-bit register or SRAM */ From ebde6f8acba92abfc203585198a54f47e83e2cd0 Mon Sep 17 00:00:00 2001 From: Thomas Jarosch Date: Mon, 16 May 2011 06:28:15 +0000 Subject: [PATCH 23/28] vmxnet3: Fix inconsistent LRO state after initialization During initialization of vmxnet3, the state of LRO gets out of sync with netdev->features. This leads to very poor TCP performance in a IP forwarding setup and is hitting many VMware users. Simplified call sequence: 1. vmxnet3_declare_features() initializes "adapter->lro" to true. 2. The kernel automatically disables LRO if IP forwarding is enabled, so vmxnet3_set_flags() gets called. This also updates netdev->features. 3. Now vmxnet3_setup_driver_shared() is called. "adapter->lro" is still set to true and LRO gets enabled again, even though netdev->features shows it's disabled. Fix it by updating "adapter->lro", too. The private vmxnet3 adapter flags are scheduled for removal in net-next, see commit a0d2730c9571aeba793cb5d3009094ee1d8fda35 "net: vmxnet3: convert to hw_features". Patch applies to 2.6.37 / 2.6.38 and 2.6.39-rc6. Please CC: comments. Signed-off-by: Thomas Jarosch Acked-by: Stephen Hemminger Signed-off-by: David S. Miller --- drivers/net/vmxnet3/vmxnet3_ethtool.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/vmxnet3/vmxnet3_ethtool.c b/drivers/net/vmxnet3/vmxnet3_ethtool.c index 51f2ef142a5b..976467253d20 100644 --- a/drivers/net/vmxnet3/vmxnet3_ethtool.c +++ b/drivers/net/vmxnet3/vmxnet3_ethtool.c @@ -311,6 +311,9 @@ vmxnet3_set_flags(struct net_device *netdev, u32 data) /* toggle the LRO feature*/ netdev->features ^= NETIF_F_LRO; + /* Update private LRO flag */ + adapter->lro = lro_requested; + /* update harware LRO capability accordingly */ if (lro_requested) adapter->shared->devRead.misc.uptFeatures |= From 6f404e441d169afc90929ef5e451ec9779c1f11a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micha=C5=82=20Miros=C5=82aw?= Date: Mon, 16 May 2011 15:14:21 -0400 Subject: [PATCH 24/28] net: Change netdev_fix_features messages loglevel MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Those reduced to DEBUG can possibly be triggered by unprivileged processes and are nothing exceptional. Illegal checksum combinations can only be caused by driver bug, so promote those messages to WARN. Since GSO without SG will now only cause DEBUG message from netdev_fix_features(), remove the workaround from register_netdevice(). Signed-off-by: Michał Mirosław Signed-off-by: David S. Miller --- net/core/dev.c | 22 ++++++++-------------- 1 file changed, 8 insertions(+), 14 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index 92009440d28b..b624fe4d9bd7 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5186,27 +5186,27 @@ u32 netdev_fix_features(struct net_device *dev, u32 features) /* Fix illegal checksum combinations */ if ((features & NETIF_F_HW_CSUM) && (features & (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))) { - netdev_info(dev, "mixed HW and IP checksum settings.\n"); + netdev_warn(dev, "mixed HW and IP checksum settings.\n"); features &= ~(NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM); } if ((features & NETIF_F_NO_CSUM) && (features & (NETIF_F_HW_CSUM|NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))) { - netdev_info(dev, "mixed no checksumming and other settings.\n"); + netdev_warn(dev, "mixed no checksumming and other settings.\n"); features &= ~(NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM|NETIF_F_HW_CSUM); } /* Fix illegal SG+CSUM combinations. */ if ((features & NETIF_F_SG) && !(features & NETIF_F_ALL_CSUM)) { - netdev_info(dev, - "Dropping NETIF_F_SG since no checksum feature.\n"); + netdev_dbg(dev, + "Dropping NETIF_F_SG since no checksum feature.\n"); features &= ~NETIF_F_SG; } /* TSO requires that SG is present as well. */ if ((features & NETIF_F_ALL_TSO) && !(features & NETIF_F_SG)) { - netdev_info(dev, "Dropping TSO features since no SG feature.\n"); + netdev_dbg(dev, "Dropping TSO features since no SG feature.\n"); features &= ~NETIF_F_ALL_TSO; } @@ -5216,7 +5216,7 @@ u32 netdev_fix_features(struct net_device *dev, u32 features) /* Software GSO depends on SG. */ if ((features & NETIF_F_GSO) && !(features & NETIF_F_SG)) { - netdev_info(dev, "Dropping NETIF_F_GSO since no SG feature.\n"); + netdev_dbg(dev, "Dropping NETIF_F_GSO since no SG feature.\n"); features &= ~NETIF_F_GSO; } @@ -5226,13 +5226,13 @@ u32 netdev_fix_features(struct net_device *dev, u32 features) if (!((features & NETIF_F_GEN_CSUM) || (features & (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM)) == (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))) { - netdev_info(dev, + netdev_dbg(dev, "Dropping NETIF_F_UFO since no checksum offload features.\n"); features &= ~NETIF_F_UFO; } if (!(features & NETIF_F_SG)) { - netdev_info(dev, + netdev_dbg(dev, "Dropping NETIF_F_UFO since no NETIF_F_SG feature.\n"); features &= ~NETIF_F_UFO; } @@ -5414,12 +5414,6 @@ int register_netdevice(struct net_device *dev) dev->features |= NETIF_F_SOFT_FEATURES; dev->wanted_features = dev->features & dev->hw_features; - /* Avoid warning from netdev_fix_features() for GSO without SG */ - if (!(dev->wanted_features & NETIF_F_SG)) { - dev->wanted_features &= ~NETIF_F_GSO; - dev->features &= ~NETIF_F_GSO; - } - /* Enable GRO and NETIF_F_HIGHDMA for vlans by default, * vlan_dev_init() will do the dev->features check, so these features * are enabled only if supported by underlying device. From 372b2312010bece1e36f577d6c99a6193ec54cbd Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 17 May 2011 13:56:59 -0400 Subject: [PATCH 25/28] net: use hlist_del_rcu() in dev_change_name() Using plain hlist_del() in dev_change_name() is wrong since a concurrent reader can crash trying to dereference LIST_POISON1. Bug introduced in commit 72c9528bab94 (net: Introduce dev_get_by_name_rcu()) Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/dev.c b/net/core/dev.c index b624fe4d9bd7..30a4078b3fa2 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1007,7 +1007,7 @@ rollback: } write_lock_bh(&dev_base_lock); - hlist_del(&dev->name_hlist); + hlist_del_rcu(&dev->name_hlist); write_unlock_bh(&dev_base_lock); synchronize_rcu(); From b9eb8b8752804cecbacdb4d24b52e823cf07f107 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Tue, 17 May 2011 15:38:57 -0400 Subject: [PATCH 26/28] net: recvmmsg: Strip MSG_WAITFORONE when calling recvmsg recvmmsg fails on a raw socket with EINVAL. The reason for this is packet_recvmsg checks the incoming flags: err = -EINVAL; if (flags & ~(MSG_PEEK|MSG_DONTWAIT|MSG_TRUNC|MSG_CMSG_COMPAT|MSG_ERRQUEUE)) goto out; This patch strips out MSG_WAITFORONE when calling recvmmsg which fixes the issue. Signed-off-by: Anton Blanchard Cc: stable@kernel.org [2.6.34+] Signed-off-by: David S. Miller --- net/socket.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/socket.c b/net/socket.c index 310d16b1b3c9..65b2310a2cb2 100644 --- a/net/socket.c +++ b/net/socket.c @@ -2122,14 +2122,16 @@ int __sys_recvmmsg(int fd, struct mmsghdr __user *mmsg, unsigned int vlen, */ if (MSG_CMSG_COMPAT & flags) { err = __sys_recvmsg(sock, (struct msghdr __user *)compat_entry, - &msg_sys, flags, datagrams); + &msg_sys, flags & ~MSG_WAITFORONE, + datagrams); if (err < 0) break; err = __put_user(err, &compat_entry->msg_len); ++compat_entry; } else { err = __sys_recvmsg(sock, (struct msghdr __user *)entry, - &msg_sys, flags, datagrams); + &msg_sys, flags & ~MSG_WAITFORONE, + datagrams); if (err < 0) break; err = put_user(err, &entry->msg_len); From 604ae14ffb6d75d6eef4757859226b758d6bf9e3 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Mon, 16 May 2011 10:37:39 +0000 Subject: [PATCH 27/28] net: Change netdev_fix_features messages loglevel Cool, how about we make 'Features changed' debug as well? This way userspace can't fill up the log just by tweaking tun features with an ioctl. Signed-off-by: Michael S. Tsirkin Signed-off-by: David S. Miller --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/dev.c b/net/core/dev.c index 30a4078b3fa2..acd742379344 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5258,7 +5258,7 @@ void netdev_update_features(struct net_device *dev) if (dev->features == features) return; - netdev_info(dev, "Features changed: 0x%08x -> 0x%08x\n", + netdev_dbg(dev, "Features changed: 0x%08x -> 0x%08x\n", dev->features, features); if (dev->netdev_ops->ndo_set_features) From 7cc31a9ae1477abc79d5992b3afe889f25c50c99 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micha=C5=82=20Miros=C5=82aw?= Date: Tue, 17 May 2011 16:50:02 -0400 Subject: [PATCH 28/28] net: ethtool: fix IPV6 checksum feature name string MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Michał Mirosław Signed-off-by: David S. Miller --- net/core/ethtool.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/ethtool.c b/net/core/ethtool.c index 74ead9eca126..f33752533119 100644 --- a/net/core/ethtool.c +++ b/net/core/ethtool.c @@ -330,7 +330,7 @@ static const char netdev_features_strings[ETHTOOL_DEV_FEATURE_WORDS * 32][ETH_GS /* NETIF_F_IP_CSUM */ "tx-checksum-ipv4", /* NETIF_F_NO_CSUM */ "tx-checksum-unneeded", /* NETIF_F_HW_CSUM */ "tx-checksum-ip-generic", - /* NETIF_F_IPV6_CSUM */ "tx_checksum-ipv6", + /* NETIF_F_IPV6_CSUM */ "tx-checksum-ipv6", /* NETIF_F_HIGHDMA */ "highdma", /* NETIF_F_FRAGLIST */ "tx-scatter-gather-fraglist", /* NETIF_F_HW_VLAN_TX */ "tx-vlan-hw-insert",