From 1b784140474e4fc94281a49e96c67d29df0efbde Mon Sep 17 00:00:00 2001 From: Ying Xue Date: Mon, 2 Mar 2015 15:37:48 +0800 Subject: net: Remove iocb argument from sendmsg and recvmsg After TIPC doesn't depend on iocb argument in its internal implementations of sendmsg() and recvmsg() hooks defined in proto structure, no any user is using iocb argument in them at all now. Then we can drop the redundant iocb argument completely from kinds of implementations of both sendmsg() and recvmsg() in the entire networking stack. Cc: Christoph Hellwig Suggested-by: Al Viro Signed-off-by: Ying Xue Signed-off-by: David S. Miller --- net/netlink/af_netlink.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'net/netlink/af_netlink.c') diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 2702673f0f23..a96025c0583f 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -2256,8 +2256,7 @@ static void netlink_cmsg_recv_pktinfo(struct msghdr *msg, struct sk_buff *skb) put_cmsg(msg, SOL_NETLINK, NETLINK_PKTINFO, sizeof(info), &info); } -static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock, - struct msghdr *msg, size_t len) +static int netlink_sendmsg(struct socket *sock, struct msghdr *msg, size_t len) { struct sock *sk = sock->sk; struct netlink_sock *nlk = nlk_sk(sk); @@ -2346,8 +2345,7 @@ out: return err; } -static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock, - struct msghdr *msg, size_t len, +static int netlink_recvmsg(struct socket *sock, struct msghdr *msg, size_t len, int flags) { struct scm_cookie scm; -- cgit v1.2.3 From b06eee59b1e547f0f401820388ca68c624298a64 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 18 Mar 2015 20:01:17 +1100 Subject: netlink: Use rhashtable max_size instead of max_shift This patch converts netlink to use rhashtable max_size instead of the obsolete max_shift. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- net/netlink/af_netlink.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/netlink/af_netlink.c') diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 6b0f21950e09..d97aed628bda 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -3123,7 +3123,7 @@ static int __init netlink_proto_init(void) .key_offset = offsetof(struct netlink_sock, portid), .key_len = sizeof(u32), /* portid */ .hashfn = jhash, - .max_shift = 16, /* 64K */ + .max_size = 65536, }; if (err != 0) -- cgit v1.2.3 From c428ecd1a21f1457ca3beb4df71b8a079c410e41 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Fri, 20 Mar 2015 21:57:01 +1100 Subject: netlink: Move namespace into hash key Currently the name space is a de facto key because it has to match before we find an object in the hash table. However, it isn't in the hash value so all objects from different name spaces with the same port ID hash to the same bucket. This is bad as the number of name spaces is unbounded. This patch fixes this by using the namespace when doing the hash. Because the namespace field doesn't lie next to the portid field in the netlink socket, this patch switches over to the rhashtable interface without a fixed key. This patch also uses the new inlined rhashtable interface where possible. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- net/netlink/af_netlink.c | 88 ++++++++++++++++++++++++++++++------------------ 1 file changed, 56 insertions(+), 32 deletions(-) (limited to 'net/netlink/af_netlink.c') diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index d97aed628bda..72c6b55af741 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -116,6 +116,8 @@ static ATOMIC_NOTIFIER_HEAD(netlink_chain); static DEFINE_SPINLOCK(netlink_tap_lock); static struct list_head netlink_tap_all __read_mostly; +static const struct rhashtable_params netlink_rhashtable_params; + static inline u32 netlink_group_mask(u32 group) { return group ? 1 << (group - 1) : 0; @@ -970,41 +972,49 @@ netlink_unlock_table(void) struct netlink_compare_arg { - struct net *net; + possible_net_t pnet; u32 portid; + char trailer[]; }; -static bool netlink_compare(void *ptr, void *arg) +#define netlink_compare_arg_len offsetof(struct netlink_compare_arg, trailer) + +static inline int netlink_compare(struct rhashtable_compare_arg *arg, + const void *ptr) { - struct netlink_compare_arg *x = arg; - struct sock *sk = ptr; + const struct netlink_compare_arg *x = arg->key; + const struct netlink_sock *nlk = ptr; - return nlk_sk(sk)->portid == x->portid && - net_eq(sock_net(sk), x->net); + return nlk->portid != x->portid || + !net_eq(sock_net(&nlk->sk), read_pnet(&x->pnet)); +} + +static void netlink_compare_arg_init(struct netlink_compare_arg *arg, + struct net *net, u32 portid) +{ + memset(arg, 0, sizeof(*arg)); + write_pnet(&arg->pnet, net); + arg->portid = portid; } static struct sock *__netlink_lookup(struct netlink_table *table, u32 portid, struct net *net) { - struct netlink_compare_arg arg = { - .net = net, - .portid = portid, - }; + struct netlink_compare_arg arg; - return rhashtable_lookup_compare(&table->hash, &portid, - &netlink_compare, &arg); + netlink_compare_arg_init(&arg, net, portid); + return rhashtable_lookup_fast(&table->hash, &arg, + netlink_rhashtable_params); } -static bool __netlink_insert(struct netlink_table *table, struct sock *sk) +static int __netlink_insert(struct netlink_table *table, struct sock *sk) { - struct netlink_compare_arg arg = { - .net = sock_net(sk), - .portid = nlk_sk(sk)->portid, - }; + struct netlink_compare_arg arg; - return rhashtable_lookup_compare_insert(&table->hash, - &nlk_sk(sk)->node, - &netlink_compare, &arg); + netlink_compare_arg_init(&arg, sock_net(sk), nlk_sk(sk)->portid); + return rhashtable_lookup_insert_key(&table->hash, &arg, + &nlk_sk(sk)->node, + netlink_rhashtable_params); } static struct sock *netlink_lookup(struct net *net, int protocol, u32 portid) @@ -1066,9 +1076,10 @@ static int netlink_insert(struct sock *sk, u32 portid) nlk_sk(sk)->portid = portid; sock_hold(sk); - err = 0; - if (!__netlink_insert(table, sk)) { - err = -EADDRINUSE; + err = __netlink_insert(table, sk); + if (err) { + if (err == -EEXIST) + err = -EADDRINUSE; sock_put(sk); } @@ -1082,7 +1093,8 @@ static void netlink_remove(struct sock *sk) struct netlink_table *table; table = &nl_table[sk->sk_protocol]; - if (rhashtable_remove(&table->hash, &nlk_sk(sk)->node)) { + if (!rhashtable_remove_fast(&table->hash, &nlk_sk(sk)->node, + netlink_rhashtable_params)) { WARN_ON(atomic_read(&sk->sk_refcnt) == 1); __sock_put(sk); } @@ -3114,17 +3126,28 @@ static struct pernet_operations __net_initdata netlink_net_ops = { .exit = netlink_net_exit, }; +static inline u32 netlink_hash(const void *data, u32 seed) +{ + const struct netlink_sock *nlk = data; + struct netlink_compare_arg arg; + + netlink_compare_arg_init(&arg, sock_net(&nlk->sk), nlk->portid); + return jhash(&arg, netlink_compare_arg_len, seed); +} + +static const struct rhashtable_params netlink_rhashtable_params = { + .head_offset = offsetof(struct netlink_sock, node), + .key_len = netlink_compare_arg_len, + .hashfn = jhash, + .obj_hashfn = netlink_hash, + .obj_cmpfn = netlink_compare, + .max_size = 65536, +}; + static int __init netlink_proto_init(void) { int i; int err = proto_register(&netlink_proto, 0); - struct rhashtable_params ht_params = { - .head_offset = offsetof(struct netlink_sock, node), - .key_offset = offsetof(struct netlink_sock, portid), - .key_len = sizeof(u32), /* portid */ - .hashfn = jhash, - .max_size = 65536, - }; if (err != 0) goto out; @@ -3136,7 +3159,8 @@ static int __init netlink_proto_init(void) goto panic; for (i = 0; i < MAX_LINKS; i++) { - if (rhashtable_init(&nl_table[i].hash, &ht_params) < 0) { + if (rhashtable_init(&nl_table[i].hash, + &netlink_rhashtable_params) < 0) { while (--i > 0) rhashtable_destroy(&nl_table[i].hash); kfree(nl_table); -- cgit v1.2.3 From 8f2ddaac302a8ca364cb04efb19cd3b5cd058910 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Sat, 21 Mar 2015 14:14:03 +1100 Subject: netlink: Remove netlink_compare_arg.trailer Instead of computing the offset from trailer, this patch computes netlink_compare_arg_len from the offset of portid and then adds 4 to it. This allows trailer to be removed. Reported-by: David Miller Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- net/netlink/af_netlink.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'net/netlink/af_netlink.c') diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 72c6b55af741..651792141f07 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -974,10 +974,11 @@ struct netlink_compare_arg { possible_net_t pnet; u32 portid; - char trailer[]; }; -#define netlink_compare_arg_len offsetof(struct netlink_compare_arg, trailer) +/* Doing sizeof directly may yield 4 extra bytes on 64-bit. */ +#define netlink_compare_arg_len \ + (offsetof(struct netlink_compare_arg, portid) + sizeof(u32)) static inline int netlink_compare(struct rhashtable_compare_arg *arg, const void *ptr) -- cgit v1.2.3 From 11b58ba146ccd7b105c4962c75f2e744053c85bc Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 24 Mar 2015 00:50:22 +1100 Subject: netlink: Use default rhashtable hashfn This patch removes the explicit jhash value for the hashfn parameter of rhashtable. As the key length is a multiple of 4, this means that we will actually end up using jhash2. Signed-off-by: Herbert Xu Acked-by: Thomas Graf Signed-off-by: David S. Miller --- net/netlink/af_netlink.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'net/netlink/af_netlink.c') diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 651792141f07..e2f7f28148e0 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -3133,13 +3133,12 @@ static inline u32 netlink_hash(const void *data, u32 seed) struct netlink_compare_arg arg; netlink_compare_arg_init(&arg, sock_net(&nlk->sk), nlk->portid); - return jhash(&arg, netlink_compare_arg_len, seed); + return jhash2((u32 *)&arg, netlink_compare_arg_len / sizeof(u32), seed); } static const struct rhashtable_params netlink_rhashtable_params = { .head_offset = offsetof(struct netlink_sock, node), .key_len = netlink_compare_arg_len, - .hashfn = jhash, .obj_hashfn = netlink_hash, .obj_cmpfn = netlink_compare, .max_size = 65536, -- cgit v1.2.3 From b5e2c150ac914f28a28833b57397bec0b0a2bd5f Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Tue, 24 Mar 2015 20:42:19 +0000 Subject: rhashtable: Disable automatic shrinking by default Introduce a new bool automatic_shrinking to require the user to explicitly opt-in to automatic shrinking of tables. Signed-off-by: Thomas Graf Signed-off-by: David S. Miller --- include/linux/rhashtable.h | 7 +++++-- lib/rhashtable.c | 2 +- net/netfilter/nft_hash.c | 1 + net/netlink/af_netlink.c | 1 + net/tipc/socket.c | 1 + 5 files changed, 9 insertions(+), 3 deletions(-) (limited to 'net/netlink/af_netlink.c') diff --git a/include/linux/rhashtable.h b/include/linux/rhashtable.h index 0e1f975ad101..ae26c494e230 100644 --- a/include/linux/rhashtable.h +++ b/include/linux/rhashtable.h @@ -2,7 +2,7 @@ * Resizable, Scalable, Concurrent Hash Table * * Copyright (c) 2015 Herbert Xu - * Copyright (c) 2014 Thomas Graf + * Copyright (c) 2014-2015 Thomas Graf * Copyright (c) 2008-2014 Patrick McHardy * * Code partially derived from nft_hash @@ -104,6 +104,7 @@ struct rhashtable; * @min_size: Minimum size while shrinking * @nulls_base: Base value to generate nulls marker * @insecure_elasticity: Set to true to disable chain length checks + * @automatic_shrinking: Enable automatic shrinking of tables * @locks_mul: Number of bucket locks to allocate per cpu (default: 128) * @hashfn: Hash function (default: jhash2 if !(key_len % 4), or jhash) * @obj_hashfn: Function to hash object @@ -118,6 +119,7 @@ struct rhashtable_params { unsigned int min_size; u32 nulls_base; bool insecure_elasticity; + bool automatic_shrinking; size_t locks_mul; rht_hashfn_t hashfn; rht_obj_hashfn_t obj_hashfn; @@ -784,7 +786,8 @@ static inline int rhashtable_remove_fast( goto out; atomic_dec(&ht->nelems); - if (rht_shrink_below_30(ht, tbl)) + if (unlikely(ht->p.automatic_shrinking && + rht_shrink_below_30(ht, tbl))) schedule_work(&ht->run_work); out: diff --git a/lib/rhashtable.c b/lib/rhashtable.c index 50abe4fec4b8..50374d181148 100644 --- a/lib/rhashtable.c +++ b/lib/rhashtable.c @@ -367,7 +367,7 @@ static void rht_deferred_worker(struct work_struct *work) if (rht_grow_above_75(ht, tbl)) rhashtable_expand(ht); - else if (rht_shrink_below_30(ht, tbl)) + else if (ht->p.automatic_shrinking && rht_shrink_below_30(ht, tbl)) rhashtable_shrink(ht); err = rhashtable_rehash_table(ht); diff --git a/net/netfilter/nft_hash.c b/net/netfilter/nft_hash.c index ad3966976cf5..8577a37af18b 100644 --- a/net/netfilter/nft_hash.c +++ b/net/netfilter/nft_hash.c @@ -172,6 +172,7 @@ static const struct rhashtable_params nft_hash_params = { .head_offset = offsetof(struct nft_hash_elem, node), .key_offset = offsetof(struct nft_hash_elem, key), .hashfn = jhash, + .automatic_shrinking = true, }; static int nft_hash_init(const struct nft_set *set, diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index e2f7f28148e0..4caa809dbbe0 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -3142,6 +3142,7 @@ static const struct rhashtable_params netlink_rhashtable_params = { .obj_hashfn = netlink_hash, .obj_cmpfn = netlink_compare, .max_size = 65536, + .automatic_shrinking = true, }; static int __init netlink_proto_init(void) diff --git a/net/tipc/socket.c b/net/tipc/socket.c index 094710519477..ee90d74d7516 100644 --- a/net/tipc/socket.c +++ b/net/tipc/socket.c @@ -2297,6 +2297,7 @@ static const struct rhashtable_params tsk_rht_params = { .key_len = sizeof(u32), /* portid */ .max_size = 1048576, .min_size = 256, + .automatic_shrinking = true, }; int tipc_sk_rht_init(struct net *net) -- cgit v1.2.3 From 49f7b33e63fec9d16e7ee62ba8f8ab4159cbdc26 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Wed, 25 Mar 2015 13:07:45 +0000 Subject: rhashtable: provide len to obj_hashfn nftables sets will be converted to use so called setextensions, moving the key to a non-fixed position. To hash it, the obj_hashfn must be used, however it so far doesn't receive the length parameter. Pass the key length to obj_hashfn() and convert existing users. Signed-off-by: Patrick McHardy Signed-off-by: Pablo Neira Ayuso --- include/linux/rhashtable.h | 6 ++++-- lib/rhashtable.c | 2 +- net/netlink/af_netlink.c | 2 +- 3 files changed, 6 insertions(+), 4 deletions(-) (limited to 'net/netlink/af_netlink.c') diff --git a/include/linux/rhashtable.h b/include/linux/rhashtable.h index 99f2e49a8a07..e23d242d1230 100644 --- a/include/linux/rhashtable.h +++ b/include/linux/rhashtable.h @@ -88,7 +88,7 @@ struct rhashtable_compare_arg { }; typedef u32 (*rht_hashfn_t)(const void *data, u32 len, u32 seed); -typedef u32 (*rht_obj_hashfn_t)(const void *data, u32 seed); +typedef u32 (*rht_obj_hashfn_t)(const void *data, u32 len, u32 seed); typedef int (*rht_obj_cmpfn_t)(struct rhashtable_compare_arg *arg, const void *obj); @@ -242,7 +242,9 @@ static inline unsigned int rht_head_hashfn( const char *ptr = rht_obj(ht, he); return likely(params.obj_hashfn) ? - rht_bucket_index(tbl, params.obj_hashfn(ptr, tbl->hash_rnd)) : + rht_bucket_index(tbl, params.obj_hashfn(ptr, params.key_len ?: + ht->p.key_len, + tbl->hash_rnd)) : rht_key_hashfn(ht, tbl, ptr + params.key_offset, params); } diff --git a/lib/rhashtable.c b/lib/rhashtable.c index 4b7b7e672b93..4898442b837f 100644 --- a/lib/rhashtable.c +++ b/lib/rhashtable.c @@ -691,7 +691,7 @@ static u32 rhashtable_jhash2(const void *key, u32 length, u32 seed) * struct rhash_head node; * }; * - * u32 my_hash_fn(const void *data, u32 seed) + * u32 my_hash_fn(const void *data, u32 len, u32 seed) * { * struct test_obj *obj = data; * diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 4caa809dbbe0..19909d0786a2 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -3127,7 +3127,7 @@ static struct pernet_operations __net_initdata netlink_net_ops = { .exit = netlink_net_exit, }; -static inline u32 netlink_hash(const void *data, u32 seed) +static inline u32 netlink_hash(const void *data, u32 len, u32 seed) { const struct netlink_sock *nlk = data; struct netlink_compare_arg arg; -- cgit v1.2.3