* Re: [PATCH bpf-next v3 0/3] support flow dissector in BPF_PROG_TEST_RUN
From: Daniel Borkmann @ 2019-01-29 0:24 UTC (permalink / raw)
To: Stanislav Fomichev, netdev; +Cc: davem, ast
In-Reply-To: <20190128165355.229403-1-sdf@google.com>
On 01/28/2019 05:53 PM, Stanislav Fomichev wrote:
> This patch series adds support for testing flow dissector BPF programs by
> extending already existing BPF_PROG_TEST_RUN. The goal is to have a
> packet as an input and `struct bpf_flow_key' as an output. That way
> we can easily test flow dissector programs' behavior.
> I've also modified existing test_progs.c test to do a simple flow
> dissector run as well.
>
> * first patch introduces new __skb_flow_bpf_dissect to simplify
> sharing between __skb_flow_bpf_dissect and BPF_PROG_TEST_RUN
> * second patch adds actual BPF_PROG_TEST_RUN support
> * third patch adds example usage to the selftests
>
> v3:
> * rebased on top of latest bpf-next
>
> v2:
> * loop over 'kattr->test.repeat' inside of
> bpf_prog_test_run_flow_dissector, don't reuse
> bpf_test_run/bpf_test_run_one
>
> Stanislav Fomichev (3):
> net/flow_dissector: move bpf case into __skb_flow_bpf_dissect
> bpf: add BPF_PROG_TEST_RUN support for flow dissector
> selftests/bpf: add simple BPF_PROG_TEST_RUN examples for flow
> dissector
>
> include/linux/bpf.h | 3 +
> include/linux/skbuff.h | 5 +
> net/bpf/test_run.c | 82 +++++++++++++++++
> net/core/filter.c | 1 +
> net/core/flow_dissector.c | 92 +++++++++++--------
> tools/testing/selftests/bpf/Makefile | 3 +
> .../selftests/bpf/flow_dissector_load.c | 43 +--------
> .../selftests/bpf/flow_dissector_load.h | 55 +++++++++++
> tools/testing/selftests/bpf/test_progs.c | 78 +++++++++++++++-
> 9 files changed, 284 insertions(+), 78 deletions(-)
> create mode 100644 tools/testing/selftests/bpf/flow_dissector_load.h
>
Applied, thanks!
^ permalink raw reply
* Re: [PATCH] bpf/core.c - silence warning messages
From: valdis.kletnieks @ 2019-01-28 23:35 UTC (permalink / raw)
To: Song Liu; +Cc: Alexei Starovoitov, Daniel Borkmann, Networking, open list
In-Reply-To: <CAPhsuW54z8ao4KOKG9ZjdXus6LTBH2qwAi8LgPFZ_qb_ObKsRg@mail.gmail.com>
On Mon, 28 Jan 2019 09:18:45 -0800, Song Liu said:
> On Sun, Jan 27, 2019 at 8:43 PM <valdis.kletnieks@vt.edu> wrote:
> > The attached patch silences the warnings, because we *know* we're overwriting
> > the default initializer. That leaves bpf/core.c with only 6 other warnings,
> > which become more visible in comparison.
>
> My concern is that this will also mute the warning for other parts of
> bpf/core.c.
I checked and there weren't any warnings for other parts of the file. Also, this message
doesn't even happen unless you build with W=1, which apparently happens so rarely
that nobody else has submitted a patch.
Is there a high likelihood that another overwrite of an initializer is going to
be included in the source?
> Maybe we should move bpf_opcode_in_insntable() to a separate file, and mute
> warning for that file?
Seems to be overkill - the intent of this patch was mostly to make the *other*
warnings issued with W=1 more noticable.
^ permalink raw reply
* Re: [PATCH 0/3] gcc-plugins: Introduce stackinit plugin
From: Alexander Popov @ 2019-01-29 0:12 UTC (permalink / raw)
To: Kees Cook, linux-kernel
Cc: Ard Biesheuvel, Laura Abbott, xen-devel, dri-devel, intel-gfx,
intel-wired-lan, netdev, linux-usb, linux-fsdevel, linux-mm, dev,
linux-kbuild, linux-security-module, kernel-hardening, Greg KH,
Jann Horn, William Kucharski, Jani Nikula, Edwin Zimmerman,
Matthew Wilcox, Jeff Kirsher
In-Reply-To: <20190123110349.35882-1-keescook@chromium.org>
On 23.01.2019 14:03, Kees Cook wrote:
> This adds a new plugin "stackinit" that attempts to perform unconditional
> initialization of all stack variables
Hello Kees! Hello everyone!
I was curious about the performance impact of the initialization of all stack
variables. So I did a very brief test with this plugin on top of 4.20.5.
hackbench on Intel Core i7-4770 showed ~0.7% slowdown.
hackbench on Kirin 620 (ARM Cortex-A53 Octa-core 1.2GHz) showed ~1.3% slowdown.
This test involves the kernel scheduler and allocator. I can't say whether they
use stack aggressively. Maybe performance tests of other subsystems (e.g.
network subsystem) can show different numbers. Did you try?
I've heard a hypothesis that the initialization of all stack variables would
pollute CPU caches, which is critical for some types of computations. Maybe some
micro-benchmarks can disprove/confirm that?
Thanks!
Best regards,
Alexander
^ permalink raw reply
* [PATCH 02/33] netfilter: nf_tables: handle nft_object lookups via rhltable
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
Instead of linear search, use rhlist interface to look up the objects.
This fixes rulesets with thousands of named objects (quota, counters and
the like).
We only use a single table for this and consider the address of the
table we're doing the lookup in as a part of the key.
This reduces restore time of a sample ruleset with ~20k named counters
from 37 seconds to 0.8 seconds.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_tables.h | 7 ++-
net/netfilter/nf_tables_api.c | 103 +++++++++++++++++++++++++++++++++-----
net/netfilter/nft_objref.c | 3 +-
3 files changed, 98 insertions(+), 15 deletions(-)
diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h
index 325d0a6b808b..45eba7d7ab38 100644
--- a/include/net/netfilter/nf_tables.h
+++ b/include/net/netfilter/nf_tables.h
@@ -1027,14 +1027,16 @@ struct nft_object_hash_key {
*
* @list: table stateful object list node
* @key: keys that identify this object
+ * @rhlhead: nft_objname_ht node
* @genmask: generation mask
* @use: number of references to this stateful object
* @handle: unique object handle
* @ops: object operations
- * @data: object data, layout depends on type
+ * @data: object data, layout depends on type
*/
struct nft_object {
struct list_head list;
+ struct rhlist_head rhlhead;
struct nft_object_hash_key key;
u32 genmask:2,
use:30;
@@ -1052,7 +1054,8 @@ static inline void *nft_obj_data(const struct nft_object *obj)
#define nft_expr_obj(expr) *((struct nft_object **)nft_expr_priv(expr))
-struct nft_object *nft_obj_lookup(const struct nft_table *table,
+struct nft_object *nft_obj_lookup(const struct net *net,
+ const struct nft_table *table,
const struct nlattr *nla, u32 objtype,
u8 genmask);
diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index 5e213941e85b..621ff834d3a4 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -37,10 +37,16 @@ enum {
NFT_VALIDATE_DO,
};
+static struct rhltable nft_objname_ht;
+
static u32 nft_chain_hash(const void *data, u32 len, u32 seed);
static u32 nft_chain_hash_obj(const void *data, u32 len, u32 seed);
static int nft_chain_hash_cmp(struct rhashtable_compare_arg *, const void *);
+static u32 nft_objname_hash(const void *data, u32 len, u32 seed);
+static u32 nft_objname_hash_obj(const void *data, u32 len, u32 seed);
+static int nft_objname_hash_cmp(struct rhashtable_compare_arg *, const void *);
+
static const struct rhashtable_params nft_chain_ht_params = {
.head_offset = offsetof(struct nft_chain, rhlhead),
.key_offset = offsetof(struct nft_chain, name),
@@ -51,6 +57,15 @@ static const struct rhashtable_params nft_chain_ht_params = {
.automatic_shrinking = true,
};
+static const struct rhashtable_params nft_objname_ht_params = {
+ .head_offset = offsetof(struct nft_object, rhlhead),
+ .key_offset = offsetof(struct nft_object, key),
+ .hashfn = nft_objname_hash,
+ .obj_hashfn = nft_objname_hash_obj,
+ .obj_cmpfn = nft_objname_hash_cmp,
+ .automatic_shrinking = true,
+};
+
static void nft_validate_state_update(struct net *net, u8 new_validate_state)
{
switch (net->nft.validate_state) {
@@ -814,6 +829,34 @@ static int nft_chain_hash_cmp(struct rhashtable_compare_arg *arg,
return strcmp(chain->name, name);
}
+static u32 nft_objname_hash(const void *data, u32 len, u32 seed)
+{
+ const struct nft_object_hash_key *k = data;
+
+ seed ^= hash_ptr(k->table, 32);
+
+ return jhash(k->name, strlen(k->name), seed);
+}
+
+static u32 nft_objname_hash_obj(const void *data, u32 len, u32 seed)
+{
+ const struct nft_object *obj = data;
+
+ return nft_objname_hash(&obj->key, 0, seed);
+}
+
+static int nft_objname_hash_cmp(struct rhashtable_compare_arg *arg,
+ const void *ptr)
+{
+ const struct nft_object_hash_key *k = arg->key;
+ const struct nft_object *obj = ptr;
+
+ if (obj->key.table != k->table)
+ return -1;
+
+ return strcmp(obj->key.name, k->name);
+}
+
static int nf_tables_newtable(struct net *net, struct sock *nlsk,
struct sk_buff *skb, const struct nlmsghdr *nlh,
const struct nlattr * const nla[],
@@ -1070,7 +1113,7 @@ nft_chain_lookup_byhandle(const struct nft_table *table, u64 handle, u8 genmask)
return ERR_PTR(-ENOENT);
}
-static bool lockdep_commit_lock_is_held(struct net *net)
+static bool lockdep_commit_lock_is_held(const struct net *net)
{
#ifdef CONFIG_PROVE_LOCKING
return lockdep_is_held(&net->nft.commit_mutex);
@@ -4386,7 +4429,8 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set,
err = -EINVAL;
goto err2;
}
- obj = nft_obj_lookup(ctx->table, nla[NFTA_SET_ELEM_OBJREF],
+ obj = nft_obj_lookup(ctx->net, ctx->table,
+ nla[NFTA_SET_ELEM_OBJREF],
set->objtype, genmask);
if (IS_ERR(obj)) {
err = PTR_ERR(obj);
@@ -4819,18 +4863,36 @@ void nft_unregister_obj(struct nft_object_type *obj_type)
}
EXPORT_SYMBOL_GPL(nft_unregister_obj);
-struct nft_object *nft_obj_lookup(const struct nft_table *table,
+struct nft_object *nft_obj_lookup(const struct net *net,
+ const struct nft_table *table,
const struct nlattr *nla, u32 objtype,
u8 genmask)
{
+ struct nft_object_hash_key k = { .table = table };
+ char search[NFT_OBJ_MAXNAMELEN];
+ struct rhlist_head *tmp, *list;
struct nft_object *obj;
- list_for_each_entry_rcu(obj, &table->objects, list) {
- if (!nla_strcmp(nla, obj->key.name) &&
- objtype == obj->ops->type->type &&
- nft_active_genmask(obj, genmask))
+ nla_strlcpy(search, nla, sizeof(search));
+ k.name = search;
+
+ WARN_ON_ONCE(!rcu_read_lock_held() &&
+ !lockdep_commit_lock_is_held(net));
+
+ rcu_read_lock();
+ list = rhltable_lookup(&nft_objname_ht, &k, nft_objname_ht_params);
+ if (!list)
+ goto out;
+
+ rhl_for_each_entry_rcu(obj, tmp, list, rhlhead) {
+ if (objtype == obj->ops->type->type &&
+ nft_active_genmask(obj, genmask)) {
+ rcu_read_unlock();
return obj;
+ }
}
+out:
+ rcu_read_unlock();
return ERR_PTR(-ENOENT);
}
EXPORT_SYMBOL_GPL(nft_obj_lookup);
@@ -4988,7 +5050,7 @@ static int nf_tables_newobj(struct net *net, struct sock *nlsk,
}
objtype = ntohl(nla_get_be32(nla[NFTA_OBJ_TYPE]));
- obj = nft_obj_lookup(table, nla[NFTA_OBJ_NAME], objtype, genmask);
+ obj = nft_obj_lookup(net, table, nla[NFTA_OBJ_NAME], objtype, genmask);
if (IS_ERR(obj)) {
err = PTR_ERR(obj);
if (err != -ENOENT) {
@@ -5027,9 +5089,18 @@ static int nf_tables_newobj(struct net *net, struct sock *nlsk,
if (err < 0)
goto err3;
+ err = rhltable_insert(&nft_objname_ht, &obj->rhlhead,
+ nft_objname_ht_params);
+ if (err < 0)
+ goto err4;
+
list_add_tail_rcu(&obj->list, &table->objects);
table->use++;
return 0;
+err4:
+ /* queued in transaction log */
+ INIT_LIST_HEAD(&obj->list);
+ return err;
err3:
kfree(obj->key.name);
err2:
@@ -5215,7 +5286,7 @@ static int nf_tables_getobj(struct net *net, struct sock *nlsk,
}
objtype = ntohl(nla_get_be32(nla[NFTA_OBJ_TYPE]));
- obj = nft_obj_lookup(table, nla[NFTA_OBJ_NAME], objtype, genmask);
+ obj = nft_obj_lookup(net, table, nla[NFTA_OBJ_NAME], objtype, genmask);
if (IS_ERR(obj)) {
NL_SET_BAD_ATTR(extack, nla[NFTA_OBJ_NAME]);
return PTR_ERR(obj);
@@ -5280,7 +5351,7 @@ static int nf_tables_delobj(struct net *net, struct sock *nlsk,
obj = nft_obj_lookup_byhandle(table, attr, objtype, genmask);
} else {
attr = nla[NFTA_OBJ_NAME];
- obj = nft_obj_lookup(table, attr, objtype, genmask);
+ obj = nft_obj_lookup(net, table, attr, objtype, genmask);
}
if (IS_ERR(obj)) {
@@ -6406,6 +6477,7 @@ static void nf_tables_commit_chain(struct net *net, struct nft_chain *chain)
static void nft_obj_del(struct nft_object *obj)
{
+ rhltable_remove(&nft_objname_ht, &obj->rhlhead, nft_objname_ht_params);
list_del_rcu(&obj->list);
}
@@ -6721,7 +6793,7 @@ static int __nf_tables_abort(struct net *net)
break;
case NFT_MSG_NEWOBJ:
trans->ctx.table->use--;
- list_del_rcu(&nft_trans_obj(trans)->list);
+ nft_obj_del(nft_trans_obj(trans));
break;
case NFT_MSG_DELOBJ:
trans->ctx.table->use++;
@@ -7397,12 +7469,18 @@ static int __init nf_tables_module_init(void)
if (err < 0)
goto err3;
+ err = rhltable_init(&nft_objname_ht, &nft_objname_ht_params);
+ if (err < 0)
+ goto err4;
+
/* must be last */
err = nfnetlink_subsys_register(&nf_tables_subsys);
if (err < 0)
- goto err4;
+ goto err5;
return err;
+err5:
+ rhltable_destroy(&nft_objname_ht);
err4:
unregister_netdevice_notifier(&nf_tables_flowtable_notifier);
err3:
@@ -7422,6 +7500,7 @@ static void __exit nf_tables_module_exit(void)
unregister_pernet_subsys(&nf_tables_net_ops);
cancel_work_sync(&trans_destroy_work);
rcu_barrier();
+ rhltable_destroy(&nft_objname_ht);
nf_tables_core_module_exit();
}
diff --git a/net/netfilter/nft_objref.c b/net/netfilter/nft_objref.c
index 58eb75ad61bf..c1f2adf198a0 100644
--- a/net/netfilter/nft_objref.c
+++ b/net/netfilter/nft_objref.c
@@ -38,7 +38,8 @@ static int nft_objref_init(const struct nft_ctx *ctx,
return -EINVAL;
objtype = ntohl(nla_get_be32(tb[NFTA_OBJREF_IMM_TYPE]));
- obj = nft_obj_lookup(ctx->table, tb[NFTA_OBJREF_IMM_NAME], objtype,
+ obj = nft_obj_lookup(ctx->net, ctx->table,
+ tb[NFTA_OBJREF_IMM_NAME], objtype,
genmask);
if (IS_ERR(obj))
return -ENOENT;
--
2.11.0
^ permalink raw reply related
* [PATCH 06/33] netfilter: nf_tables: Support RULE_ID reference in new rule
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Phil Sutter <phil@nwl.cc>
To allow for a batch to contain rules in arbitrary ordering, introduce
NFTA_RULE_POSITION_ID attribute which works just like NFTA_RULE_POSITION
but contains the ID of another rule within the same batch. This helps
iptables-nft-restore handling dumps with mixed insert/append commands
correctly.
Note that NFTA_RULE_POSITION takes precedence over
NFTA_RULE_POSITION_ID, so if the former is present, the latter is
ignored.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/uapi/linux/netfilter/nf_tables.h | 2 ++
net/netfilter/nf_tables_api.c | 9 +++++++++
2 files changed, 11 insertions(+)
diff --git a/include/uapi/linux/netfilter/nf_tables.h b/include/uapi/linux/netfilter/nf_tables.h
index 7de4f1bdaf06..99ca95b830b6 100644
--- a/include/uapi/linux/netfilter/nf_tables.h
+++ b/include/uapi/linux/netfilter/nf_tables.h
@@ -219,6 +219,7 @@ enum nft_chain_attributes {
* @NFTA_RULE_POSITION: numeric handle of the previous rule (NLA_U64)
* @NFTA_RULE_USERDATA: user data (NLA_BINARY, NFT_USERDATA_MAXLEN)
* @NFTA_RULE_ID: uniquely identifies a rule in a transaction (NLA_U32)
+ * @NFTA_RULE_POSITION_ID: transaction unique identifier of the previous rule (NLA_U32)
*/
enum nft_rule_attributes {
NFTA_RULE_UNSPEC,
@@ -231,6 +232,7 @@ enum nft_rule_attributes {
NFTA_RULE_USERDATA,
NFTA_RULE_PAD,
NFTA_RULE_ID,
+ NFTA_RULE_POSITION_ID,
__NFTA_RULE_MAX
};
#define NFTA_RULE_MAX (__NFTA_RULE_MAX - 1)
diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index 621ff834d3a4..d88c86c5b433 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -2610,6 +2610,9 @@ static int nft_table_validate(struct net *net, const struct nft_table *table)
return 0;
}
+static struct nft_rule *nft_rule_lookup_byid(const struct net *net,
+ const struct nlattr *nla);
+
#define NFT_RULE_MAXEXPRS 128
static int nf_tables_newrule(struct net *net, struct sock *nlsk,
@@ -2679,6 +2682,12 @@ static int nf_tables_newrule(struct net *net, struct sock *nlsk,
NL_SET_BAD_ATTR(extack, nla[NFTA_RULE_POSITION]);
return PTR_ERR(old_rule);
}
+ } else if (nla[NFTA_RULE_POSITION_ID]) {
+ old_rule = nft_rule_lookup_byid(net, nla[NFTA_RULE_POSITION_ID]);
+ if (IS_ERR(old_rule)) {
+ NL_SET_BAD_ATTR(extack, nla[NFTA_RULE_POSITION_ID]);
+ return PTR_ERR(old_rule);
+ }
}
}
--
2.11.0
^ permalink raw reply related
* [PATCH 05/33] netfilter: physdev: relax br_netfilter dependency
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
Following command:
iptables -D FORWARD -m physdev ...
causes connectivity loss in some setups.
Reason is that iptables userspace will probe kernel for the module revision
of the physdev patch, and physdev has an artificial dependency on
br_netfilter (xt_physdev use makes no sense unless a br_netfilter module
is loaded).
This causes the "phydev" module to be loaded, which in turn enables the
"call-iptables" infrastructure.
bridged packets might then get dropped by the iptables ruleset.
The better fix would be to change the "call-iptables" defaults to 0 and
enforce explicit setting to 1, but that breaks backwards compatibility.
This does the next best thing: add a request_module call to checkentry.
This was a stray '-D ... -m physdev' won't activate br_netfilter
anymore.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/br_netfilter.h | 1 -
net/bridge/br_netfilter_hooks.c | 5 -----
net/netfilter/xt_physdev.c | 9 +++++++--
3 files changed, 7 insertions(+), 8 deletions(-)
diff --git a/include/net/netfilter/br_netfilter.h b/include/net/netfilter/br_netfilter.h
index 4cd56808ac4e..89808ce293c4 100644
--- a/include/net/netfilter/br_netfilter.h
+++ b/include/net/netfilter/br_netfilter.h
@@ -43,7 +43,6 @@ static inline struct rtable *bridge_parent_rtable(const struct net_device *dev)
}
struct net_device *setup_pre_routing(struct sk_buff *skb);
-void br_netfilter_enable(void);
#if IS_ENABLED(CONFIG_IPV6)
int br_validate_ipv6(struct net *net, struct sk_buff *skb);
diff --git a/net/bridge/br_netfilter_hooks.c b/net/bridge/br_netfilter_hooks.c
index c93c35bb73dd..40d058378b52 100644
--- a/net/bridge/br_netfilter_hooks.c
+++ b/net/bridge/br_netfilter_hooks.c
@@ -881,11 +881,6 @@ static const struct nf_br_ops br_ops = {
.br_dev_xmit_hook = br_nf_dev_xmit,
};
-void br_netfilter_enable(void)
-{
-}
-EXPORT_SYMBOL_GPL(br_netfilter_enable);
-
/* For br_nf_post_routing, we need (prio = NF_BR_PRI_LAST), because
* br_dev_queue_push_xmit is called afterwards */
static const struct nf_hook_ops br_nf_ops[] = {
diff --git a/net/netfilter/xt_physdev.c b/net/netfilter/xt_physdev.c
index 4034d70bff39..b2e39cb6a590 100644
--- a/net/netfilter/xt_physdev.c
+++ b/net/netfilter/xt_physdev.c
@@ -96,8 +96,7 @@ physdev_mt(const struct sk_buff *skb, struct xt_action_param *par)
static int physdev_mt_check(const struct xt_mtchk_param *par)
{
const struct xt_physdev_info *info = par->matchinfo;
-
- br_netfilter_enable();
+ static bool brnf_probed __read_mostly;
if (!(info->bitmask & XT_PHYSDEV_OP_MASK) ||
info->bitmask & ~XT_PHYSDEV_OP_MASK)
@@ -111,6 +110,12 @@ static int physdev_mt_check(const struct xt_mtchk_param *par)
if (par->hook_mask & (1 << NF_INET_LOCAL_OUT))
return -EINVAL;
}
+
+ if (!brnf_probed) {
+ brnf_probed = true;
+ request_module("br_netfilter");
+ }
+
return 0;
}
--
2.11.0
^ permalink raw reply related
* [PATCH 04/33] netfilter: conntrack: remove helper hook again
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
place them into the confirm one.
Old:
hook (300): ipv4/6_help() first call helper, then seqadj.
hook (INT_MAX): confirm
Now:
hook (INT_MAX): confirm, first call helper, then seqadj, then confirm
Not having the extra call is noticeable in bechmarks.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
net/netfilter/nf_conntrack_proto.c | 142 ++++++++++---------------------------
1 file changed, 36 insertions(+), 106 deletions(-)
diff --git a/net/netfilter/nf_conntrack_proto.c b/net/netfilter/nf_conntrack_proto.c
index 859f5d07a915..2bbc32d939e4 100644
--- a/net/netfilter/nf_conntrack_proto.c
+++ b/net/netfilter/nf_conntrack_proto.c
@@ -364,55 +364,55 @@ void nf_ct_l4proto_pernet_unregister(struct net *net,
}
EXPORT_SYMBOL_GPL(nf_ct_l4proto_pernet_unregister);
-static unsigned int ipv4_helper(void *priv,
- struct sk_buff *skb,
- const struct nf_hook_state *state)
+static unsigned int nf_confirm(struct sk_buff *skb,
+ unsigned int protoff,
+ struct nf_conn *ct,
+ enum ip_conntrack_info ctinfo)
{
- struct nf_conn *ct;
- enum ip_conntrack_info ctinfo;
const struct nf_conn_help *help;
- const struct nf_conntrack_helper *helper;
-
- /* This is where we call the helper: as the packet goes out. */
- ct = nf_ct_get(skb, &ctinfo);
- if (!ct || ctinfo == IP_CT_RELATED_REPLY)
- return NF_ACCEPT;
help = nfct_help(ct);
- if (!help)
- return NF_ACCEPT;
+ if (help) {
+ const struct nf_conntrack_helper *helper;
+ int ret;
+
+ /* rcu_read_lock()ed by nf_hook_thresh */
+ helper = rcu_dereference(help->helper);
+ if (helper) {
+ ret = helper->help(skb,
+ protoff,
+ ct, ctinfo);
+ if (ret != NF_ACCEPT)
+ return ret;
+ }
+ }
- /* rcu_read_lock()ed by nf_hook_thresh */
- helper = rcu_dereference(help->helper);
- if (!helper)
- return NF_ACCEPT;
+ if (test_bit(IPS_SEQ_ADJUST_BIT, &ct->status) &&
+ !nf_is_loopback_packet(skb)) {
+ if (!nf_ct_seq_adjust(skb, ct, ctinfo, protoff)) {
+ NF_CT_STAT_INC_ATOMIC(nf_ct_net(ct), drop);
+ return NF_DROP;
+ }
+ }
- return helper->help(skb, skb_network_offset(skb) + ip_hdrlen(skb),
- ct, ctinfo);
+ /* We've seen it coming out the other side: confirm it */
+ return nf_conntrack_confirm(skb);
}
static unsigned int ipv4_confirm(void *priv,
struct sk_buff *skb,
const struct nf_hook_state *state)
{
- struct nf_conn *ct;
enum ip_conntrack_info ctinfo;
+ struct nf_conn *ct;
ct = nf_ct_get(skb, &ctinfo);
if (!ct || ctinfo == IP_CT_RELATED_REPLY)
- goto out;
+ return nf_conntrack_confirm(skb);
- /* adjust seqs for loopback traffic only in outgoing direction */
- if (test_bit(IPS_SEQ_ADJUST_BIT, &ct->status) &&
- !nf_is_loopback_packet(skb)) {
- if (!nf_ct_seq_adjust(skb, ct, ctinfo, ip_hdrlen(skb))) {
- NF_CT_STAT_INC_ATOMIC(nf_ct_net(ct), drop);
- return NF_DROP;
- }
- }
-out:
- /* We've seen it coming out the other side: confirm it */
- return nf_conntrack_confirm(skb);
+ return nf_confirm(skb,
+ skb_network_offset(skb) + ip_hdrlen(skb),
+ ct, ctinfo);
}
static unsigned int ipv4_conntrack_in(void *priv,
@@ -461,24 +461,12 @@ static const struct nf_hook_ops ipv4_conntrack_ops[] = {
.priority = NF_IP_PRI_CONNTRACK,
},
{
- .hook = ipv4_helper,
- .pf = NFPROTO_IPV4,
- .hooknum = NF_INET_POST_ROUTING,
- .priority = NF_IP_PRI_CONNTRACK_HELPER,
- },
- {
.hook = ipv4_confirm,
.pf = NFPROTO_IPV4,
.hooknum = NF_INET_POST_ROUTING,
.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
},
{
- .hook = ipv4_helper,
- .pf = NFPROTO_IPV4,
- .hooknum = NF_INET_LOCAL_IN,
- .priority = NF_IP_PRI_CONNTRACK_HELPER,
- },
- {
.hook = ipv4_confirm,
.pf = NFPROTO_IPV4,
.hooknum = NF_INET_LOCAL_IN,
@@ -623,31 +611,21 @@ static unsigned int ipv6_confirm(void *priv,
struct nf_conn *ct;
enum ip_conntrack_info ctinfo;
unsigned char pnum = ipv6_hdr(skb)->nexthdr;
- int protoff;
__be16 frag_off;
+ int protoff;
ct = nf_ct_get(skb, &ctinfo);
if (!ct || ctinfo == IP_CT_RELATED_REPLY)
- goto out;
+ return nf_conntrack_confirm(skb);
protoff = ipv6_skip_exthdr(skb, sizeof(struct ipv6hdr), &pnum,
&frag_off);
if (protoff < 0 || (frag_off & htons(~0x7)) != 0) {
pr_debug("proto header not found\n");
- goto out;
+ return nf_conntrack_confirm(skb);
}
- /* adjust seqs for loopback traffic only in outgoing direction */
- if (test_bit(IPS_SEQ_ADJUST_BIT, &ct->status) &&
- !nf_is_loopback_packet(skb)) {
- if (!nf_ct_seq_adjust(skb, ct, ctinfo, protoff)) {
- NF_CT_STAT_INC_ATOMIC(nf_ct_net(ct), drop);
- return NF_DROP;
- }
- }
-out:
- /* We've seen it coming out the other side: confirm it */
- return nf_conntrack_confirm(skb);
+ return nf_confirm(skb, protoff, ct, ctinfo);
}
static unsigned int ipv6_conntrack_in(void *priv,
@@ -664,42 +642,6 @@ static unsigned int ipv6_conntrack_local(void *priv,
return nf_conntrack_in(skb, state);
}
-static unsigned int ipv6_helper(void *priv,
- struct sk_buff *skb,
- const struct nf_hook_state *state)
-{
- struct nf_conn *ct;
- const struct nf_conn_help *help;
- const struct nf_conntrack_helper *helper;
- enum ip_conntrack_info ctinfo;
- __be16 frag_off;
- int protoff;
- u8 nexthdr;
-
- /* This is where we call the helper: as the packet goes out. */
- ct = nf_ct_get(skb, &ctinfo);
- if (!ct || ctinfo == IP_CT_RELATED_REPLY)
- return NF_ACCEPT;
-
- help = nfct_help(ct);
- if (!help)
- return NF_ACCEPT;
- /* rcu_read_lock()ed by nf_hook_thresh */
- helper = rcu_dereference(help->helper);
- if (!helper)
- return NF_ACCEPT;
-
- nexthdr = ipv6_hdr(skb)->nexthdr;
- protoff = ipv6_skip_exthdr(skb, sizeof(struct ipv6hdr), &nexthdr,
- &frag_off);
- if (protoff < 0 || (frag_off & htons(~0x7)) != 0) {
- pr_debug("proto header not found\n");
- return NF_ACCEPT;
- }
-
- return helper->help(skb, protoff, ct, ctinfo);
-}
-
static const struct nf_hook_ops ipv6_conntrack_ops[] = {
{
.hook = ipv6_conntrack_in,
@@ -714,24 +656,12 @@ static const struct nf_hook_ops ipv6_conntrack_ops[] = {
.priority = NF_IP6_PRI_CONNTRACK,
},
{
- .hook = ipv6_helper,
- .pf = NFPROTO_IPV6,
- .hooknum = NF_INET_POST_ROUTING,
- .priority = NF_IP6_PRI_CONNTRACK_HELPER,
- },
- {
.hook = ipv6_confirm,
.pf = NFPROTO_IPV6,
.hooknum = NF_INET_POST_ROUTING,
.priority = NF_IP6_PRI_LAST,
},
{
- .hook = ipv6_helper,
- .pf = NFPROTO_IPV6,
- .hooknum = NF_INET_LOCAL_IN,
- .priority = NF_IP6_PRI_CONNTRACK_HELPER,
- },
- {
.hook = ipv6_confirm,
.pf = NFPROTO_IPV6,
.hooknum = NF_INET_LOCAL_IN,
--
2.11.0
^ permalink raw reply related
* [PATCH 08/33] netfilter: conntrack: handle icmp pkt_to_tuple helper via direct calls
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
rather than handling them via indirect call, use a direct one instead.
This leaves GRE as the last user of this indirect call facility.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_conntrack_l4proto.h | 10 ++++++++++
net/netfilter/nf_conntrack_core.c | 6 ++++++
net/netfilter/nf_conntrack_proto_icmp.c | 5 ++---
net/netfilter/nf_conntrack_proto_icmpv6.c | 9 ++++-----
4 files changed, 22 insertions(+), 8 deletions(-)
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index 6ca4355ae3dc..46d554806eb3 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -90,6 +90,16 @@ struct nf_conntrack_l4proto {
struct module *me;
};
+bool icmp_pkt_to_tuple(const struct sk_buff *skb,
+ unsigned int dataoff,
+ struct net *net,
+ struct nf_conntrack_tuple *tuple);
+
+bool icmpv6_pkt_to_tuple(const struct sk_buff *skb,
+ unsigned int dataoff,
+ struct net *net,
+ struct nf_conntrack_tuple *tuple);
+
int nf_conntrack_icmpv4_error(struct nf_conn *tmpl,
struct sk_buff *skb,
unsigned int dataoff,
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 5dfd95ae510d..b3840d36c3a6 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -274,6 +274,12 @@ nf_ct_get_tuple(const struct sk_buff *skb,
tuple->dst.protonum = protonum;
tuple->dst.dir = IP_CT_DIR_ORIGINAL;
+ switch (protonum) {
+ case IPPROTO_ICMPV6:
+ return icmpv6_pkt_to_tuple(skb, dataoff, net, tuple);
+ case IPPROTO_ICMP:
+ return icmp_pkt_to_tuple(skb, dataoff, net, tuple);
+ }
if (unlikely(l4proto->pkt_to_tuple))
return l4proto->pkt_to_tuple(skb, dataoff, net, tuple);
diff --git a/net/netfilter/nf_conntrack_proto_icmp.c b/net/netfilter/nf_conntrack_proto_icmp.c
index 381a7f3ea68a..805c1fe5b837 100644
--- a/net/netfilter/nf_conntrack_proto_icmp.c
+++ b/net/netfilter/nf_conntrack_proto_icmp.c
@@ -25,8 +25,8 @@
static const unsigned int nf_ct_icmp_timeout = 30*HZ;
-static bool icmp_pkt_to_tuple(const struct sk_buff *skb, unsigned int dataoff,
- struct net *net, struct nf_conntrack_tuple *tuple)
+bool icmp_pkt_to_tuple(const struct sk_buff *skb, unsigned int dataoff,
+ struct net *net, struct nf_conntrack_tuple *tuple)
{
const struct icmphdr *hp;
struct icmphdr _hdr;
@@ -347,7 +347,6 @@ static struct nf_proto_net *icmp_get_net_proto(struct net *net)
const struct nf_conntrack_l4proto nf_conntrack_l4proto_icmp =
{
.l4proto = IPPROTO_ICMP,
- .pkt_to_tuple = icmp_pkt_to_tuple,
.invert_tuple = icmp_invert_tuple,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.tuple_to_nlattr = icmp_tuple_to_nlattr,
diff --git a/net/netfilter/nf_conntrack_proto_icmpv6.c b/net/netfilter/nf_conntrack_proto_icmpv6.c
index 93f27d0bcb75..20cd55e55e41 100644
--- a/net/netfilter/nf_conntrack_proto_icmpv6.c
+++ b/net/netfilter/nf_conntrack_proto_icmpv6.c
@@ -30,10 +30,10 @@
static const unsigned int nf_ct_icmpv6_timeout = 30*HZ;
-static bool icmpv6_pkt_to_tuple(const struct sk_buff *skb,
- unsigned int dataoff,
- struct net *net,
- struct nf_conntrack_tuple *tuple)
+bool icmpv6_pkt_to_tuple(const struct sk_buff *skb,
+ unsigned int dataoff,
+ struct net *net,
+ struct nf_conntrack_tuple *tuple)
{
const struct icmp6hdr *hp;
struct icmp6hdr _hdr;
@@ -358,7 +358,6 @@ static struct nf_proto_net *icmpv6_get_net_proto(struct net *net)
const struct nf_conntrack_l4proto nf_conntrack_l4proto_icmpv6 =
{
.l4proto = IPPROTO_ICMPV6,
- .pkt_to_tuple = icmpv6_pkt_to_tuple,
.invert_tuple = icmpv6_invert_tuple,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.tuple_to_nlattr = icmpv6_tuple_to_nlattr,
--
2.11.0
^ permalink raw reply related
* [PATCH 09/33] netfilter: conntrack: gre: convert rwlock to rcu
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
We can use gre. Lock is only needed when a new expectation is added.
In case a single spinlock proves to be problematic we can either add one
per netns or use an array of locks combined with net_hash_mix() or similar
to pick the 'correct' one.
But given this is only needed for an expectation rather than per packet
a single one should be ok.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/linux/netfilter/nf_conntrack_proto_gre.h | 1 +
net/netfilter/nf_conntrack_proto_gre.c | 37 ++++++++++--------------
2 files changed, 16 insertions(+), 22 deletions(-)
diff --git a/include/linux/netfilter/nf_conntrack_proto_gre.h b/include/linux/netfilter/nf_conntrack_proto_gre.h
index 6989e2e4eabf..222c9d3d453f 100644
--- a/include/linux/netfilter/nf_conntrack_proto_gre.h
+++ b/include/linux/netfilter/nf_conntrack_proto_gre.h
@@ -19,6 +19,7 @@ struct nf_conn;
struct nf_ct_gre_keymap {
struct list_head list;
struct nf_conntrack_tuple tuple;
+ struct rcu_head rcu;
};
enum grep_conntrack {
diff --git a/net/netfilter/nf_conntrack_proto_gre.c b/net/netfilter/nf_conntrack_proto_gre.c
index 8899b51aad44..34dd89485be2 100644
--- a/net/netfilter/nf_conntrack_proto_gre.c
+++ b/net/netfilter/nf_conntrack_proto_gre.c
@@ -49,6 +49,8 @@ static const unsigned int gre_timeouts[GRE_CT_MAX] = {
};
static unsigned int proto_gre_net_id __read_mostly;
+/* used when expectation is added */
+static DEFINE_SPINLOCK(keymap_lock);
static inline struct netns_proto_gre *gre_pernet(struct net *net)
{
@@ -60,12 +62,12 @@ static void nf_ct_gre_keymap_flush(struct net *net)
struct netns_proto_gre *net_gre = gre_pernet(net);
struct nf_ct_gre_keymap *km, *tmp;
- write_lock_bh(&net_gre->keymap_lock);
+ spin_lock_bh(&keymap_lock);
list_for_each_entry_safe(km, tmp, &net_gre->keymap_list, list) {
- list_del(&km->list);
- kfree(km);
+ list_del_rcu(&km->list);
+ kfree_rcu(km, rcu);
}
- write_unlock_bh(&net_gre->keymap_lock);
+ spin_unlock_bh(&keymap_lock);
}
static inline int gre_key_cmpfn(const struct nf_ct_gre_keymap *km,
@@ -85,14 +87,12 @@ static __be16 gre_keymap_lookup(struct net *net, struct nf_conntrack_tuple *t)
struct nf_ct_gre_keymap *km;
__be16 key = 0;
- read_lock_bh(&net_gre->keymap_lock);
- list_for_each_entry(km, &net_gre->keymap_list, list) {
+ list_for_each_entry_rcu(km, &net_gre->keymap_list, list) {
if (gre_key_cmpfn(km, t)) {
key = km->tuple.src.u.gre.key;
break;
}
}
- read_unlock_bh(&net_gre->keymap_lock);
pr_debug("lookup src key 0x%x for ", key);
nf_ct_dump_tuple(t);
@@ -112,14 +112,10 @@ int nf_ct_gre_keymap_add(struct nf_conn *ct, enum ip_conntrack_dir dir,
kmp = &ct_pptp_info->keymap[dir];
if (*kmp) {
/* check whether it's a retransmission */
- read_lock_bh(&net_gre->keymap_lock);
- list_for_each_entry(km, &net_gre->keymap_list, list) {
- if (gre_key_cmpfn(km, t) && km == *kmp) {
- read_unlock_bh(&net_gre->keymap_lock);
+ list_for_each_entry_rcu(km, &net_gre->keymap_list, list) {
+ if (gre_key_cmpfn(km, t) && km == *kmp)
return 0;
- }
}
- read_unlock_bh(&net_gre->keymap_lock);
pr_debug("trying to override keymap_%s for ct %p\n",
dir == IP_CT_DIR_REPLY ? "reply" : "orig", ct);
return -EEXIST;
@@ -134,9 +130,9 @@ int nf_ct_gre_keymap_add(struct nf_conn *ct, enum ip_conntrack_dir dir,
pr_debug("adding new entry %p: ", km);
nf_ct_dump_tuple(&km->tuple);
- write_lock_bh(&net_gre->keymap_lock);
+ spin_lock_bh(&keymap_lock);
list_add_tail(&km->list, &net_gre->keymap_list);
- write_unlock_bh(&net_gre->keymap_lock);
+ spin_unlock_bh(&keymap_lock);
return 0;
}
@@ -145,24 +141,22 @@ EXPORT_SYMBOL_GPL(nf_ct_gre_keymap_add);
/* destroy the keymap entries associated with specified master ct */
void nf_ct_gre_keymap_destroy(struct nf_conn *ct)
{
- struct net *net = nf_ct_net(ct);
- struct netns_proto_gre *net_gre = gre_pernet(net);
struct nf_ct_pptp_master *ct_pptp_info = nfct_help_data(ct);
enum ip_conntrack_dir dir;
pr_debug("entering for ct %p\n", ct);
- write_lock_bh(&net_gre->keymap_lock);
+ spin_lock_bh(&keymap_lock);
for (dir = IP_CT_DIR_ORIGINAL; dir < IP_CT_DIR_MAX; dir++) {
if (ct_pptp_info->keymap[dir]) {
pr_debug("removing %p from list\n",
ct_pptp_info->keymap[dir]);
- list_del(&ct_pptp_info->keymap[dir]->list);
- kfree(ct_pptp_info->keymap[dir]);
+ list_del_rcu(&ct_pptp_info->keymap[dir]->list);
+ kfree_rcu(ct_pptp_info->keymap[dir], rcu);
ct_pptp_info->keymap[dir] = NULL;
}
}
- write_unlock_bh(&net_gre->keymap_lock);
+ spin_unlock_bh(&keymap_lock);
}
EXPORT_SYMBOL_GPL(nf_ct_gre_keymap_destroy);
@@ -365,7 +359,6 @@ static int gre_init_net(struct net *net)
struct nf_proto_net *nf = &net_gre->nf;
int i;
- rwlock_init(&net_gre->keymap_lock);
INIT_LIST_HEAD(&net_gre->keymap_list);
for (i = 0; i < GRE_CT_MAX; i++)
net_gre->gre_timeouts[i] = gre_timeouts[i];
--
2.11.0
^ permalink raw reply related
* [PATCH 10/33] netfilter: conntrack: gre: switch module to be built-in
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
This makes the last of the modular l4 trackers 'bool'.
After this, all infrastructure to handle dynamic l4 protocol registration
becomes obsolete and can be removed in followup patches.
Old:
302824 net/netfilter/nf_conntrack.ko
21504 net/netfilter/nf_conntrack_proto_gre.ko
New:
313728 net/netfilter/nf_conntrack.ko
Old:
text data bss dec hex filename
6281 1732 4 8017 1f51 nf_conntrack_proto_gre.ko
108356 20613 236 129205 1f8b5 nf_conntrack.ko
New:
112095 21381 240 133716 20a54 nf_conntrack.ko
The size increase is only temporary.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/linux/netfilter/nf_conntrack_proto_gre.h | 14 +---
include/net/netfilter/ipv4/nf_conntrack_ipv4.h | 3 +
include/net/netfilter/nf_conntrack_l4proto.h | 7 ++
include/net/netns/conntrack.h | 17 +++++
net/netfilter/Kconfig | 2 +-
net/netfilter/Makefile | 3 +-
net/netfilter/nf_conntrack_proto.c | 7 +-
net/netfilter/nf_conntrack_proto_gre.c | 93 +++++-------------------
net/netfilter/nfnetlink_cttimeout.c | 7 +-
9 files changed, 55 insertions(+), 98 deletions(-)
diff --git a/include/linux/netfilter/nf_conntrack_proto_gre.h b/include/linux/netfilter/nf_conntrack_proto_gre.h
index 222c9d3d453f..59714e9ee4ef 100644
--- a/include/linux/netfilter/nf_conntrack_proto_gre.h
+++ b/include/linux/netfilter/nf_conntrack_proto_gre.h
@@ -22,23 +22,11 @@ struct nf_ct_gre_keymap {
struct rcu_head rcu;
};
-enum grep_conntrack {
- GRE_CT_UNREPLIED,
- GRE_CT_REPLIED,
- GRE_CT_MAX
-};
-
-struct netns_proto_gre {
- struct nf_proto_net nf;
- rwlock_t keymap_lock;
- struct list_head keymap_list;
- unsigned int gre_timeouts[GRE_CT_MAX];
-};
-
/* add new tuple->key_reply pair to keymap */
int nf_ct_gre_keymap_add(struct nf_conn *ct, enum ip_conntrack_dir dir,
struct nf_conntrack_tuple *t);
+void nf_ct_gre_keymap_flush(struct net *net);
/* delete keymap entries */
void nf_ct_gre_keymap_destroy(struct nf_conn *ct);
diff --git a/include/net/netfilter/ipv4/nf_conntrack_ipv4.h b/include/net/netfilter/ipv4/nf_conntrack_ipv4.h
index 135ee702c7b0..2c8c2b023848 100644
--- a/include/net/netfilter/ipv4/nf_conntrack_ipv4.h
+++ b/include/net/netfilter/ipv4/nf_conntrack_ipv4.h
@@ -22,5 +22,8 @@ extern const struct nf_conntrack_l4proto nf_conntrack_l4proto_sctp;
#ifdef CONFIG_NF_CT_PROTO_UDPLITE
extern const struct nf_conntrack_l4proto nf_conntrack_l4proto_udplite;
#endif
+#ifdef CONFIG_NF_CT_PROTO_GRE
+extern const struct nf_conntrack_l4proto nf_conntrack_l4proto_gre;
+#endif
#endif /*_NF_CONNTRACK_IPV4_H*/
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index 46d554806eb3..fded3f164dcc 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -239,4 +239,11 @@ static inline struct nf_sctp_net *nf_sctp_pernet(struct net *net)
}
#endif
+#ifdef CONFIG_NF_CT_PROTO_GRE
+static inline struct nf_gre_net *nf_gre_pernet(struct net *net)
+{
+ return &net->ct.nf_ct_proto.gre;
+}
+#endif
+
#endif /*_NF_CONNTRACK_PROTOCOL_H*/
diff --git a/include/net/netns/conntrack.h b/include/net/netns/conntrack.h
index 51cba0b8adf5..c72f413a2d4d 100644
--- a/include/net/netns/conntrack.h
+++ b/include/net/netns/conntrack.h
@@ -70,6 +70,20 @@ struct nf_sctp_net {
};
#endif
+#ifdef CONFIG_NF_CT_PROTO_GRE
+enum gre_conntrack {
+ GRE_CT_UNREPLIED,
+ GRE_CT_REPLIED,
+ GRE_CT_MAX
+};
+
+struct nf_gre_net {
+ struct nf_proto_net nf;
+ struct list_head keymap_list;
+ unsigned int timeouts[GRE_CT_MAX];
+};
+#endif
+
struct nf_ip_net {
struct nf_generic_net generic;
struct nf_tcp_net tcp;
@@ -82,6 +96,9 @@ struct nf_ip_net {
#ifdef CONFIG_NF_CT_PROTO_SCTP
struct nf_sctp_net sctp;
#endif
+#ifdef CONFIG_NF_CT_PROTO_GRE
+ struct nf_gre_net gre;
+#endif
};
struct ct_pcpu {
diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig
index beb3a69ce1d4..fefd63a243f2 100644
--- a/net/netfilter/Kconfig
+++ b/net/netfilter/Kconfig
@@ -174,7 +174,7 @@ config NF_CT_PROTO_DCCP
If unsure, say Y.
config NF_CT_PROTO_GRE
- tristate
+ bool
config NF_CT_PROTO_SCTP
bool 'SCTP protocol connection tracking support'
diff --git a/net/netfilter/Makefile b/net/netfilter/Makefile
index 1ae65a314d7a..e66067befa42 100644
--- a/net/netfilter/Makefile
+++ b/net/netfilter/Makefile
@@ -13,6 +13,7 @@ nf_conntrack-$(CONFIG_NF_CONNTRACK_EVENTS) += nf_conntrack_ecache.o
nf_conntrack-$(CONFIG_NF_CONNTRACK_LABELS) += nf_conntrack_labels.o
nf_conntrack-$(CONFIG_NF_CT_PROTO_DCCP) += nf_conntrack_proto_dccp.o
nf_conntrack-$(CONFIG_NF_CT_PROTO_SCTP) += nf_conntrack_proto_sctp.o
+nf_conntrack-$(CONFIG_NF_CT_PROTO_GRE) += nf_conntrack_proto_gre.o
obj-$(CONFIG_NETFILTER) = netfilter.o
@@ -25,8 +26,6 @@ obj-$(CONFIG_NETFILTER_NETLINK_OSF) += nfnetlink_osf.o
# connection tracking
obj-$(CONFIG_NF_CONNTRACK) += nf_conntrack.o
-obj-$(CONFIG_NF_CT_PROTO_GRE) += nf_conntrack_proto_gre.o
-
# netlink interface for nf_conntrack
obj-$(CONFIG_NF_CT_NETLINK) += nf_conntrack_netlink.o
obj-$(CONFIG_NF_CT_NETLINK_TIMEOUT) += nfnetlink_cttimeout.o
diff --git a/net/netfilter/nf_conntrack_proto.c b/net/netfilter/nf_conntrack_proto.c
index 2bbc32d939e4..e113bb2dc88d 100644
--- a/net/netfilter/nf_conntrack_proto.c
+++ b/net/netfilter/nf_conntrack_proto.c
@@ -817,6 +817,9 @@ static const struct nf_conntrack_l4proto * const builtin_l4proto[] = {
#ifdef CONFIG_NF_CT_PROTO_UDPLITE
&nf_conntrack_l4proto_udplite,
#endif
+#ifdef CONFIG_NF_CT_PROTO_GRE
+ &nf_conntrack_l4proto_gre,
+#endif
#if IS_ENABLED(CONFIG_IPV6)
&nf_conntrack_l4proto_icmpv6,
#endif /* CONFIG_IPV6 */
@@ -897,9 +900,11 @@ void nf_conntrack_proto_pernet_fini(struct net *net)
ARRAY_SIZE(builtin_l4proto));
pn->users--;
nf_ct_l4proto_unregister_sysctl(pn);
+#ifdef CONFIG_NF_CT_PROTO_GRE
+ nf_ct_gre_keymap_flush(net);
+#endif
}
-
module_param_call(hashsize, nf_conntrack_set_hashsize, param_get_uint,
&nf_conntrack_htable_size, 0600);
diff --git a/net/netfilter/nf_conntrack_proto_gre.c b/net/netfilter/nf_conntrack_proto_gre.c
index 34dd89485be2..68f9bfb79c4e 100644
--- a/net/netfilter/nf_conntrack_proto_gre.c
+++ b/net/netfilter/nf_conntrack_proto_gre.c
@@ -48,18 +48,17 @@ static const unsigned int gre_timeouts[GRE_CT_MAX] = {
[GRE_CT_REPLIED] = 180*HZ,
};
-static unsigned int proto_gre_net_id __read_mostly;
/* used when expectation is added */
static DEFINE_SPINLOCK(keymap_lock);
-static inline struct netns_proto_gre *gre_pernet(struct net *net)
+static inline struct nf_gre_net *gre_pernet(struct net *net)
{
- return net_generic(net, proto_gre_net_id);
+ return &net->ct.nf_ct_proto.gre;
}
-static void nf_ct_gre_keymap_flush(struct net *net)
+void nf_ct_gre_keymap_flush(struct net *net)
{
- struct netns_proto_gre *net_gre = gre_pernet(net);
+ struct nf_gre_net *net_gre = gre_pernet(net);
struct nf_ct_gre_keymap *km, *tmp;
spin_lock_bh(&keymap_lock);
@@ -83,7 +82,7 @@ static inline int gre_key_cmpfn(const struct nf_ct_gre_keymap *km,
/* look up the source key for a given tuple */
static __be16 gre_keymap_lookup(struct net *net, struct nf_conntrack_tuple *t)
{
- struct netns_proto_gre *net_gre = gre_pernet(net);
+ struct nf_gre_net *net_gre = gre_pernet(net);
struct nf_ct_gre_keymap *km;
__be16 key = 0;
@@ -105,7 +104,7 @@ int nf_ct_gre_keymap_add(struct nf_conn *ct, enum ip_conntrack_dir dir,
struct nf_conntrack_tuple *t)
{
struct net *net = nf_ct_net(ct);
- struct netns_proto_gre *net_gre = gre_pernet(net);
+ struct nf_gre_net *net_gre = gre_pernet(net);
struct nf_ct_pptp_master *ct_pptp_info = nfct_help_data(ct);
struct nf_ct_gre_keymap **kmp, *km;
@@ -210,7 +209,7 @@ static void gre_print_conntrack(struct seq_file *s, struct nf_conn *ct)
static unsigned int *gre_get_timeouts(struct net *net)
{
- return gre_pernet(net)->gre_timeouts;
+ return gre_pernet(net)->timeouts;
}
/* Returns verdict for packet, and may modify conntrack */
@@ -272,13 +271,13 @@ static int gre_timeout_nlattr_to_obj(struct nlattr *tb[],
struct net *net, void *data)
{
unsigned int *timeouts = data;
- struct netns_proto_gre *net_gre = gre_pernet(net);
+ struct nf_gre_net *net_gre = gre_pernet(net);
if (!timeouts)
timeouts = gre_get_timeouts(net);
/* set default timeouts for GRE. */
- timeouts[GRE_CT_UNREPLIED] = net_gre->gre_timeouts[GRE_CT_UNREPLIED];
- timeouts[GRE_CT_REPLIED] = net_gre->gre_timeouts[GRE_CT_REPLIED];
+ timeouts[GRE_CT_UNREPLIED] = net_gre->timeouts[GRE_CT_UNREPLIED];
+ timeouts[GRE_CT_REPLIED] = net_gre->timeouts[GRE_CT_REPLIED];
if (tb[CTA_TIMEOUT_GRE_UNREPLIED]) {
timeouts[GRE_CT_UNREPLIED] =
@@ -332,10 +331,11 @@ static struct ctl_table gre_sysctl_table[] = {
};
#endif
-static int gre_kmemdup_sysctl_table(struct net *net, struct nf_proto_net *nf,
- struct netns_proto_gre *net_gre)
+static int gre_kmemdup_sysctl_table(struct net *net)
{
#ifdef CONFIG_SYSCTL
+ struct nf_gre_net *net_gre = gre_pernet(net);
+ struct nf_proto_net *nf = &net_gre->nf;
int i;
if (nf->ctl_table)
@@ -348,26 +348,25 @@ static int gre_kmemdup_sysctl_table(struct net *net, struct nf_proto_net *nf,
return -ENOMEM;
for (i = 0; i < GRE_CT_MAX; i++)
- nf->ctl_table[i].data = &net_gre->gre_timeouts[i];
+ nf->ctl_table[i].data = &net_gre->timeouts[i];
#endif
return 0;
}
static int gre_init_net(struct net *net)
{
- struct netns_proto_gre *net_gre = gre_pernet(net);
- struct nf_proto_net *nf = &net_gre->nf;
+ struct nf_gre_net *net_gre = gre_pernet(net);
int i;
INIT_LIST_HEAD(&net_gre->keymap_list);
for (i = 0; i < GRE_CT_MAX; i++)
- net_gre->gre_timeouts[i] = gre_timeouts[i];
+ net_gre->timeouts[i] = gre_timeouts[i];
- return gre_kmemdup_sysctl_table(net, nf, net_gre);
+ return gre_kmemdup_sysctl_table(net);
}
/* protocol helper struct */
-static const struct nf_conntrack_l4proto nf_conntrack_l4proto_gre4 = {
+const struct nf_conntrack_l4proto nf_conntrack_l4proto_gre = {
.l4proto = IPPROTO_GRE,
.pkt_to_tuple = gre_pkt_to_tuple,
#ifdef CONFIG_NF_CONNTRACK_PROCFS
@@ -391,61 +390,5 @@ static const struct nf_conntrack_l4proto nf_conntrack_l4proto_gre4 = {
.nla_policy = gre_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .net_id = &proto_gre_net_id,
.init_net = gre_init_net,
};
-
-static int proto_gre_net_init(struct net *net)
-{
- int ret = 0;
-
- ret = nf_ct_l4proto_pernet_register_one(net,
- &nf_conntrack_l4proto_gre4);
- if (ret < 0)
- pr_err("nf_conntrack_gre4: pernet registration failed.\n");
- return ret;
-}
-
-static void proto_gre_net_exit(struct net *net)
-{
- nf_ct_l4proto_pernet_unregister_one(net, &nf_conntrack_l4proto_gre4);
- nf_ct_gre_keymap_flush(net);
-}
-
-static struct pernet_operations proto_gre_net_ops = {
- .init = proto_gre_net_init,
- .exit = proto_gre_net_exit,
- .id = &proto_gre_net_id,
- .size = sizeof(struct netns_proto_gre),
-};
-
-static int __init nf_ct_proto_gre_init(void)
-{
- int ret;
-
- BUILD_BUG_ON(offsetof(struct netns_proto_gre, nf) != 0);
-
- ret = register_pernet_subsys(&proto_gre_net_ops);
- if (ret < 0)
- goto out_pernet;
- ret = nf_ct_l4proto_register_one(&nf_conntrack_l4proto_gre4);
- if (ret < 0)
- goto out_gre4;
-
- return 0;
-out_gre4:
- unregister_pernet_subsys(&proto_gre_net_ops);
-out_pernet:
- return ret;
-}
-
-static void __exit nf_ct_proto_gre_fini(void)
-{
- nf_ct_l4proto_unregister_one(&nf_conntrack_l4proto_gre4);
- unregister_pernet_subsys(&proto_gre_net_ops);
-}
-
-module_init(nf_ct_proto_gre_init);
-module_exit(nf_ct_proto_gre_fini);
-
-MODULE_LICENSE("GPL");
diff --git a/net/netfilter/nfnetlink_cttimeout.c b/net/netfilter/nfnetlink_cttimeout.c
index 109b0d27345a..0e3e1a018206 100644
--- a/net/netfilter/nfnetlink_cttimeout.c
+++ b/net/netfilter/nfnetlink_cttimeout.c
@@ -474,12 +474,7 @@ static int cttimeout_default_get(struct net *net, struct sock *ctnl,
break;
case IPPROTO_GRE:
#ifdef CONFIG_NF_CT_PROTO_GRE
- if (l4proto->net_id) {
- struct netns_proto_gre *net_gre;
-
- net_gre = net_generic(net, *l4proto->net_id);
- timeouts = net_gre->gre_timeouts;
- }
+ timeouts = nf_gre_pernet(net)->timeouts;
#endif
break;
case 255:
--
2.11.0
^ permalink raw reply related
* [PATCH 12/33] netfilter: conntrack: remove pkt_to_tuple callback
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
GRE is now builtin, so we can handle it via direct call and
remove the callback.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/linux/netfilter/nf_conntrack_proto_gre.h | 2 ++
include/net/netfilter/nf_conntrack_l4proto.h | 5 -----
net/netfilter/nf_conntrack_core.c | 6 ++++--
net/netfilter/nf_conntrack_proto_generic.c | 11 -----------
net/netfilter/nf_conntrack_proto_gre.c | 5 ++---
5 files changed, 8 insertions(+), 21 deletions(-)
diff --git a/include/linux/netfilter/nf_conntrack_proto_gre.h b/include/linux/netfilter/nf_conntrack_proto_gre.h
index 59714e9ee4ef..25f9a770fb84 100644
--- a/include/linux/netfilter/nf_conntrack_proto_gre.h
+++ b/include/linux/netfilter/nf_conntrack_proto_gre.h
@@ -30,5 +30,7 @@ void nf_ct_gre_keymap_flush(struct net *net);
/* delete keymap entries */
void nf_ct_gre_keymap_destroy(struct nf_conn *ct);
+bool gre_pkt_to_tuple(const struct sk_buff *skb, unsigned int dataoff,
+ struct net *net, struct nf_conntrack_tuple *tuple);
#endif /* __KERNEL__ */
#endif /* _CONNTRACK_PROTO_GRE_H */
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index 3585f8666fc0..0d4b0398aeb9 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -27,11 +27,6 @@ struct nf_conntrack_l4proto {
/* protoinfo nlattr size, closes a hole */
u16 nlattr_size;
- /* Try to fill in the third arg: dataoff is offset past network protocol
- hdr. Return true if possible. */
- bool (*pkt_to_tuple)(const struct sk_buff *skb, unsigned int dataoff,
- struct net *net, struct nf_conntrack_tuple *tuple);
-
/* Invert the per-proto part of the tuple: ie. turn xmit into reply.
* Only used by icmp, most protocols use a generic version.
*/
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index b3840d36c3a6..b71e271f2b44 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -279,9 +279,11 @@ nf_ct_get_tuple(const struct sk_buff *skb,
return icmpv6_pkt_to_tuple(skb, dataoff, net, tuple);
case IPPROTO_ICMP:
return icmp_pkt_to_tuple(skb, dataoff, net, tuple);
+#ifdef CONFIG_NF_CT_PROTO_GRE
+ case IPPROTO_GRE:
+ return gre_pkt_to_tuple(skb, dataoff, net, tuple);
+#endif
}
- if (unlikely(l4proto->pkt_to_tuple))
- return l4proto->pkt_to_tuple(skb, dataoff, net, tuple);
/* Actually only need first 4 bytes to get ports. */
inet_hdr = skb_header_pointer(skb, dataoff, sizeof(_inet_hdr), &_inet_hdr);
diff --git a/net/netfilter/nf_conntrack_proto_generic.c b/net/netfilter/nf_conntrack_proto_generic.c
index 5da19d5fbc76..5a5bf7cb6508 100644
--- a/net/netfilter/nf_conntrack_proto_generic.c
+++ b/net/netfilter/nf_conntrack_proto_generic.c
@@ -27,16 +27,6 @@ static bool nf_generic_should_process(u8 proto)
}
}
-static bool generic_pkt_to_tuple(const struct sk_buff *skb,
- unsigned int dataoff,
- struct net *net, struct nf_conntrack_tuple *tuple)
-{
- tuple->src.u.all = 0;
- tuple->dst.u.all = 0;
-
- return true;
-}
-
/* Returns verdict for packet, or -1 for invalid. */
static int generic_packet(struct nf_conn *ct,
struct sk_buff *skb,
@@ -149,7 +139,6 @@ static struct nf_proto_net *generic_get_net_proto(struct net *net)
const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic =
{
.l4proto = 255,
- .pkt_to_tuple = generic_pkt_to_tuple,
.packet = generic_packet,
#ifdef CONFIG_NF_CONNTRACK_TIMEOUT
.ctnl_timeout = {
diff --git a/net/netfilter/nf_conntrack_proto_gre.c b/net/netfilter/nf_conntrack_proto_gre.c
index 68f9bfb79c4e..04bc982b274d 100644
--- a/net/netfilter/nf_conntrack_proto_gre.c
+++ b/net/netfilter/nf_conntrack_proto_gre.c
@@ -162,8 +162,8 @@ EXPORT_SYMBOL_GPL(nf_ct_gre_keymap_destroy);
/* PUBLIC CONNTRACK PROTO HELPER FUNCTIONS */
/* gre hdr info to tuple */
-static bool gre_pkt_to_tuple(const struct sk_buff *skb, unsigned int dataoff,
- struct net *net, struct nf_conntrack_tuple *tuple)
+bool gre_pkt_to_tuple(const struct sk_buff *skb, unsigned int dataoff,
+ struct net *net, struct nf_conntrack_tuple *tuple)
{
const struct pptp_gre_header *pgrehdr;
struct pptp_gre_header _pgrehdr;
@@ -368,7 +368,6 @@ static int gre_init_net(struct net *net)
/* protocol helper struct */
const struct nf_conntrack_l4proto nf_conntrack_l4proto_gre = {
.l4proto = IPPROTO_GRE,
- .pkt_to_tuple = gre_pkt_to_tuple,
#ifdef CONFIG_NF_CONNTRACK_PROCFS
.print_conntrack = gre_print_conntrack,
#endif
--
2.11.0
^ permalink raw reply related
* [PATCH 11/33] netfilter: conntrack: remove net_id
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
No users anymore.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_conntrack_l4proto.h | 2 +-
net/netfilter/nf_conntrack_proto.c | 8 ++------
2 files changed, 3 insertions(+), 7 deletions(-)
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index fded3f164dcc..3585f8666fc0 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -79,7 +79,7 @@ struct nf_conntrack_l4proto {
/* Print out the private part of the conntrack. */
void (*print_conntrack)(struct seq_file *s, struct nf_conn *);
#endif
- unsigned int *net_id;
+
/* Init l4proto pernet data */
int (*init_net)(struct net *net);
diff --git a/net/netfilter/nf_conntrack_proto.c b/net/netfilter/nf_conntrack_proto.c
index e113bb2dc88d..47b9294a86f7 100644
--- a/net/netfilter/nf_conntrack_proto.c
+++ b/net/netfilter/nf_conntrack_proto.c
@@ -163,13 +163,9 @@ static int kill_l4proto(struct nf_conn *i, void *data)
static struct nf_proto_net *nf_ct_l4proto_net(struct net *net,
const struct nf_conntrack_l4proto *l4proto)
{
- if (l4proto->get_net_proto) {
- /* statically built-in protocols use static per-net */
+ if (l4proto->get_net_proto)
return l4proto->get_net_proto(net);
- } else if (l4proto->net_id) {
- /* ... and loadable protocols use dynamic per-net */
- return net_generic(net, *l4proto->net_id);
- }
+
return NULL;
}
--
2.11.0
^ permalink raw reply related
* [PATCH 14/33] netfilter: conntrack: remove module owner field
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
No need to get/put module owner reference, none of these can be removed
anymore.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_conntrack_l4proto.h | 4 ----
net/netfilter/nf_conntrack_proto.c | 8 --------
net/netfilter/nf_conntrack_proto_gre.c | 1 -
net/netfilter/nfnetlink_cttimeout.c | 6 ------
net/netfilter/nft_ct.c | 2 --
5 files changed, 21 deletions(-)
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index 6cec8337e684..fd8954429e87 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -74,9 +74,6 @@ struct nf_conntrack_l4proto {
/* Return the per-net protocol part. */
struct nf_proto_net *(*get_net_proto)(struct net *net);
-
- /* Module (if any) which this is connected to. */
- struct module *me;
};
bool icmp_pkt_to_tuple(const struct sk_buff *skb,
@@ -148,7 +145,6 @@ extern const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic;
const struct nf_conntrack_l4proto *__nf_ct_l4proto_find(u8 l4proto);
const struct nf_conntrack_l4proto *nf_ct_l4proto_find_get(u8 l4proto);
-void nf_ct_l4proto_put(const struct nf_conntrack_l4proto *p);
/* Protocol pernet registration. */
int nf_ct_l4proto_pernet_register_one(struct net *net,
diff --git a/net/netfilter/nf_conntrack_proto.c b/net/netfilter/nf_conntrack_proto.c
index 47b9294a86f7..3455f993cc68 100644
--- a/net/netfilter/nf_conntrack_proto.c
+++ b/net/netfilter/nf_conntrack_proto.c
@@ -139,20 +139,12 @@ const struct nf_conntrack_l4proto *nf_ct_l4proto_find_get(u8 l4num)
rcu_read_lock();
p = __nf_ct_l4proto_find(l4num);
- if (!try_module_get(p->me))
- p = &nf_conntrack_l4proto_generic;
rcu_read_unlock();
return p;
}
EXPORT_SYMBOL_GPL(nf_ct_l4proto_find_get);
-void nf_ct_l4proto_put(const struct nf_conntrack_l4proto *p)
-{
- module_put(p->me);
-}
-EXPORT_SYMBOL_GPL(nf_ct_l4proto_put);
-
static int kill_l4proto(struct nf_conn *i, void *data)
{
const struct nf_conntrack_l4proto *l4proto;
diff --git a/net/netfilter/nf_conntrack_proto_gre.c b/net/netfilter/nf_conntrack_proto_gre.c
index 04bc982b274d..0c26cfd5d29f 100644
--- a/net/netfilter/nf_conntrack_proto_gre.c
+++ b/net/netfilter/nf_conntrack_proto_gre.c
@@ -373,7 +373,6 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_gre = {
#endif
.packet = gre_packet,
.destroy = gre_destroy,
- .me = THIS_MODULE,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.tuple_to_nlattr = nf_ct_port_tuple_to_nlattr,
.nlattr_tuple_size = nf_ct_port_nlattr_tuple_size,
diff --git a/net/netfilter/nfnetlink_cttimeout.c b/net/netfilter/nfnetlink_cttimeout.c
index 0e3e1a018206..37b4f84ac153 100644
--- a/net/netfilter/nfnetlink_cttimeout.c
+++ b/net/netfilter/nfnetlink_cttimeout.c
@@ -152,7 +152,6 @@ static int cttimeout_new_timeout(struct net *net, struct sock *ctnl,
err:
kfree(timeout);
err_proto_put:
- nf_ct_l4proto_put(l4proto);
return ret;
}
@@ -302,7 +301,6 @@ static int ctnl_timeout_try_del(struct net *net, struct ctnl_timeout *timeout)
if (refcount_dec_if_one(&timeout->refcnt)) {
/* We are protected by nfnl mutex. */
list_del_rcu(&timeout->head);
- nf_ct_l4proto_put(timeout->timeout.l4proto);
nf_ct_untimeout(net, &timeout->timeout);
kfree_rcu(timeout, rcu_head);
} else {
@@ -372,10 +370,8 @@ static int cttimeout_default_set(struct net *net, struct sock *ctnl,
if (ret < 0)
goto err;
- nf_ct_l4proto_put(l4proto);
return 0;
err:
- nf_ct_l4proto_put(l4proto);
return ret;
}
@@ -511,7 +507,6 @@ static int cttimeout_default_get(struct net *net, struct sock *ctnl,
/* this avoids a loop in nfnetlink. */
return ret == -EAGAIN ? -ENOBUFS : ret;
err:
- nf_ct_l4proto_put(l4proto);
return err;
}
@@ -592,7 +587,6 @@ static void __net_exit cttimeout_net_exit(struct net *net)
list_for_each_entry_safe(cur, tmp, &net->nfct_timeout_list, head) {
list_del_rcu(&cur->head);
- nf_ct_l4proto_put(cur->timeout.l4proto);
if (refcount_dec_and_test(&cur->refcnt))
kfree_rcu(cur, rcu_head);
diff --git a/net/netfilter/nft_ct.c b/net/netfilter/nft_ct.c
index 586627c361df..3249cc059048 100644
--- a/net/netfilter/nft_ct.c
+++ b/net/netfilter/nft_ct.c
@@ -902,7 +902,6 @@ static int nft_ct_timeout_obj_init(const struct nft_ctx *ctx,
err_free_timeout:
kfree(timeout);
err_proto_put:
- nf_ct_l4proto_put(l4proto);
return ret;
}
@@ -913,7 +912,6 @@ static void nft_ct_timeout_obj_destroy(const struct nft_ctx *ctx,
struct nf_ct_timeout *timeout = priv->timeout;
nf_ct_untimeout(ctx->net, timeout);
- nf_ct_l4proto_put(timeout->l4proto);
nf_ct_netns_put(ctx->net, ctx->family);
kfree(priv->timeout);
}
--
2.11.0
^ permalink raw reply related
* [PATCH 07/33] netfilter: conntrack: handle builtin l4proto packet functions via direct calls
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
The l4 protocol trackers are invoked via indirect call: l4proto->packet().
With one exception (gre), all l4trackers are builtin, so we can make
.packet optional and use a direct call for most protocols.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_conntrack_l4proto.h | 37 +++++++++++++++++++++++
net/netfilter/nf_conntrack_core.c | 45 +++++++++++++++++++++++++++-
net/netfilter/nf_conntrack_proto_dccp.c | 8 ++---
net/netfilter/nf_conntrack_proto_icmp.c | 12 +++-----
net/netfilter/nf_conntrack_proto_icmpv6.c | 10 +++----
net/netfilter/nf_conntrack_proto_sctp.c | 12 ++++----
net/netfilter/nf_conntrack_proto_tcp.c | 11 ++++---
net/netfilter/nf_conntrack_proto_udp.c | 22 +++++++-------
8 files changed, 113 insertions(+), 44 deletions(-)
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index ae7b86f587f2..6ca4355ae3dc 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -99,6 +99,43 @@ int nf_conntrack_icmpv6_error(struct nf_conn *tmpl,
struct sk_buff *skb,
unsigned int dataoff,
const struct nf_hook_state *state);
+
+int nf_conntrack_icmp_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state);
+
+int nf_conntrack_icmpv6_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state);
+
+int nf_conntrack_udp_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state);
+int nf_conntrack_udplite_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state);
+int nf_conntrack_tcp_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state);
+int nf_conntrack_dccp_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state);
+int nf_conntrack_sctp_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state);
+
/* Existing built-in generic protocol */
extern const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic;
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 741b533148ba..5dfd95ae510d 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -1522,6 +1522,45 @@ nf_conntrack_handle_icmp(struct nf_conn *tmpl,
return ret;
}
+/* Returns verdict for packet, or -1 for invalid. */
+static int nf_conntrack_handle_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state)
+{
+ switch (nf_ct_protonum(ct)) {
+ case IPPROTO_TCP:
+ return nf_conntrack_tcp_packet(ct, skb, dataoff,
+ ctinfo, state);
+ case IPPROTO_UDP:
+ return nf_conntrack_udp_packet(ct, skb, dataoff,
+ ctinfo, state);
+ case IPPROTO_ICMP:
+ return nf_conntrack_icmp_packet(ct, skb, ctinfo, state);
+ case IPPROTO_ICMPV6:
+ return nf_conntrack_icmpv6_packet(ct, skb, ctinfo, state);
+#ifdef CONFIG_NF_CT_PROTO_UDPLITE
+ case IPPROTO_UDPLITE:
+ return nf_conntrack_udplite_packet(ct, skb, dataoff,
+ ctinfo, state);
+#endif
+#ifdef CONFIG_NF_CT_PROTO_SCTP
+ case IPPROTO_SCTP:
+ return nf_conntrack_sctp_packet(ct, skb, dataoff,
+ ctinfo, state);
+#endif
+#ifdef CONFIG_NF_CT_PROTO_DCCP
+ case IPPROTO_DCCP:
+ return nf_conntrack_dccp_packet(ct, skb, dataoff,
+ ctinfo, state);
+#endif
+ }
+
+ WARN_ON_ONCE(1);
+ return -NF_ACCEPT;
+}
+
unsigned int
nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
{
@@ -1583,7 +1622,11 @@ nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
goto out;
}
- ret = l4proto->packet(ct, skb, dataoff, ctinfo, state);
+ if (l4proto->packet)
+ ret = l4proto->packet(ct, skb, dataoff, ctinfo, state);
+ else
+ ret = nf_conntrack_handle_packet(ct, skb, dataoff, ctinfo, state);
+
if (ret <= 0) {
/* Invalid: inverse of the return code tells
* the netfilter core what to do */
diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c
index 023c1445bc39..84408ec80624 100644
--- a/net/netfilter/nf_conntrack_proto_dccp.c
+++ b/net/netfilter/nf_conntrack_proto_dccp.c
@@ -472,9 +472,10 @@ static bool dccp_error(const struct dccp_hdr *dh,
return true;
}
-static int dccp_packet(struct nf_conn *ct, struct sk_buff *skb,
- unsigned int dataoff, enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state)
+int nf_conntrack_dccp_packet(struct nf_conn *ct, struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state)
{
enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
struct dccp_hdr _dh, *dh;
@@ -839,7 +840,6 @@ static struct nf_proto_net *dccp_get_net_proto(struct net *net)
const struct nf_conntrack_l4proto nf_conntrack_l4proto_dccp = {
.l4proto = IPPROTO_DCCP,
- .packet = dccp_packet,
.can_early_drop = dccp_can_early_drop,
#ifdef CONFIG_NF_CONNTRACK_PROCFS
.print_conntrack = dccp_print_conntrack,
diff --git a/net/netfilter/nf_conntrack_proto_icmp.c b/net/netfilter/nf_conntrack_proto_icmp.c
index de64d8a5fdfd..381a7f3ea68a 100644
--- a/net/netfilter/nf_conntrack_proto_icmp.c
+++ b/net/netfilter/nf_conntrack_proto_icmp.c
@@ -68,11 +68,10 @@ static bool icmp_invert_tuple(struct nf_conntrack_tuple *tuple,
}
/* Returns verdict for packet, or -1 for invalid. */
-static int icmp_packet(struct nf_conn *ct,
- struct sk_buff *skb,
- unsigned int dataoff,
- enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state)
+int nf_conntrack_icmp_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state)
{
/* Do not immediately delete the connection after the first
successful reply to avoid excessive conntrackd traffic
@@ -350,9 +349,6 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_icmp =
.l4proto = IPPROTO_ICMP,
.pkt_to_tuple = icmp_pkt_to_tuple,
.invert_tuple = icmp_invert_tuple,
- .packet = icmp_packet,
- .destroy = NULL,
- .me = NULL,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.tuple_to_nlattr = icmp_tuple_to_nlattr,
.nlattr_tuple_size = icmp_nlattr_tuple_size,
diff --git a/net/netfilter/nf_conntrack_proto_icmpv6.c b/net/netfilter/nf_conntrack_proto_icmpv6.c
index a15eefb8e317..93f27d0bcb75 100644
--- a/net/netfilter/nf_conntrack_proto_icmpv6.c
+++ b/net/netfilter/nf_conntrack_proto_icmpv6.c
@@ -86,11 +86,10 @@ static unsigned int *icmpv6_get_timeouts(struct net *net)
}
/* Returns verdict for packet, or -1 for invalid. */
-static int icmpv6_packet(struct nf_conn *ct,
- struct sk_buff *skb,
- unsigned int dataoff,
- enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state)
+int nf_conntrack_icmpv6_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state)
{
unsigned int *timeout = nf_ct_timeout_lookup(ct);
static const u8 valid_new[] = {
@@ -361,7 +360,6 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_icmpv6 =
.l4proto = IPPROTO_ICMPV6,
.pkt_to_tuple = icmpv6_pkt_to_tuple,
.invert_tuple = icmpv6_invert_tuple,
- .packet = icmpv6_packet,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.tuple_to_nlattr = icmpv6_tuple_to_nlattr,
.nlattr_tuple_size = icmpv6_nlattr_tuple_size,
diff --git a/net/netfilter/nf_conntrack_proto_sctp.c b/net/netfilter/nf_conntrack_proto_sctp.c
index d53e3e78f605..952f2cc0fd74 100644
--- a/net/netfilter/nf_conntrack_proto_sctp.c
+++ b/net/netfilter/nf_conntrack_proto_sctp.c
@@ -357,11 +357,11 @@ static bool sctp_error(struct sk_buff *skb,
}
/* Returns verdict for packet, or -NF_ACCEPT for invalid. */
-static int sctp_packet(struct nf_conn *ct,
- struct sk_buff *skb,
- unsigned int dataoff,
- enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state)
+int nf_conntrack_sctp_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state)
{
enum sctp_conntrack new_state, old_state;
enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
@@ -759,9 +759,7 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_sctp = {
#ifdef CONFIG_NF_CONNTRACK_PROCFS
.print_conntrack = sctp_print_conntrack,
#endif
- .packet = sctp_packet,
.can_early_drop = sctp_can_early_drop,
- .me = THIS_MODULE,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.nlattr_size = SCTP_NLATTR_SIZE,
.to_nlattr = sctp_to_nlattr,
diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c
index 4dcbd51a8e97..32aac9a835d4 100644
--- a/net/netfilter/nf_conntrack_proto_tcp.c
+++ b/net/netfilter/nf_conntrack_proto_tcp.c
@@ -829,11 +829,11 @@ static noinline bool tcp_new(struct nf_conn *ct, const struct sk_buff *skb,
}
/* Returns verdict for packet, or -1 for invalid. */
-static int tcp_packet(struct nf_conn *ct,
- struct sk_buff *skb,
- unsigned int dataoff,
- enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state)
+int nf_conntrack_tcp_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state)
{
struct net *net = nf_ct_net(ct);
struct nf_tcp_net *tn = nf_tcp_pernet(net);
@@ -1535,7 +1535,6 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_tcp =
#ifdef CONFIG_NF_CONNTRACK_PROCFS
.print_conntrack = tcp_print_conntrack,
#endif
- .packet = tcp_packet,
.can_early_drop = tcp_can_early_drop,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.to_nlattr = tcp_to_nlattr,
diff --git a/net/netfilter/nf_conntrack_proto_udp.c b/net/netfilter/nf_conntrack_proto_udp.c
index b4f5d5e82031..e1f409aa0006 100644
--- a/net/netfilter/nf_conntrack_proto_udp.c
+++ b/net/netfilter/nf_conntrack_proto_udp.c
@@ -85,11 +85,11 @@ static bool udp_error(struct sk_buff *skb,
}
/* Returns verdict for packet, and may modify conntracktype */
-static int udp_packet(struct nf_conn *ct,
- struct sk_buff *skb,
- unsigned int dataoff,
- enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state)
+int nf_conntrack_udp_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state)
{
unsigned int *timeouts;
@@ -177,11 +177,11 @@ static bool udplite_error(struct sk_buff *skb,
}
/* Returns verdict for packet, and may modify conntracktype */
-static int udplite_packet(struct nf_conn *ct,
- struct sk_buff *skb,
- unsigned int dataoff,
- enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state)
+int nf_conntrack_udplite_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state)
{
unsigned int *timeouts;
@@ -319,7 +319,6 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_udp =
{
.l4proto = IPPROTO_UDP,
.allow_clash = true,
- .packet = udp_packet,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.tuple_to_nlattr = nf_ct_port_tuple_to_nlattr,
.nlattr_to_tuple = nf_ct_port_nlattr_to_tuple,
@@ -344,7 +343,6 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_udplite =
{
.l4proto = IPPROTO_UDPLITE,
.allow_clash = true,
- .packet = udplite_packet,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.tuple_to_nlattr = nf_ct_port_tuple_to_nlattr,
.nlattr_to_tuple = nf_ct_port_nlattr_to_tuple,
--
2.11.0
^ permalink raw reply related
* [PATCH 15/33] netfilter: conntrack: remove remaining l4proto indirect packet calls
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
Now that all l4trackers are builtin, no need to use a mix of direct and
indirect calls.
This removes the last two users: gre and the generic l4 protocol
tracker.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_conntrack_l4proto.h | 12 ++++------
net/netfilter/nf_conntrack_core.c | 26 +++++++++++++++------
net/netfilter/nf_conntrack_proto_generic.c | 35 ----------------------------
net/netfilter/nf_conntrack_proto_gre.c | 11 ++++-----
4 files changed, 29 insertions(+), 55 deletions(-)
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index fd8954429e87..dda028996559 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -27,13 +27,6 @@ struct nf_conntrack_l4proto {
/* protoinfo nlattr size, closes a hole */
u16 nlattr_size;
- /* Returns verdict for packet, or -1 for invalid. */
- int (*packet)(struct nf_conn *ct,
- struct sk_buff *skb,
- unsigned int dataoff,
- enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state);
-
/* Called when a conntrack entry is destroyed */
void (*destroy)(struct nf_conn *ct);
@@ -136,6 +129,11 @@ int nf_conntrack_sctp_packet(struct nf_conn *ct,
unsigned int dataoff,
enum ip_conntrack_info ctinfo,
const struct nf_hook_state *state);
+int nf_conntrack_gre_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state);
/* Existing built-in generic protocol */
extern const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic;
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index d56cb0fc82b6..264074f04615 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -1534,6 +1534,18 @@ nf_conntrack_handle_icmp(struct nf_conn *tmpl,
return ret;
}
+static int generic_packet(struct nf_conn *ct, struct sk_buff *skb,
+ enum ip_conntrack_info ctinfo)
+{
+ const unsigned int *timeout = nf_ct_timeout_lookup(ct);
+
+ if (!timeout)
+ timeout = &nf_generic_pernet(nf_ct_net(ct))->timeout;
+
+ nf_ct_refresh_acct(ct, ctinfo, skb, *timeout);
+ return NF_ACCEPT;
+}
+
/* Returns verdict for packet, or -1 for invalid. */
static int nf_conntrack_handle_packet(struct nf_conn *ct,
struct sk_buff *skb,
@@ -1567,10 +1579,14 @@ static int nf_conntrack_handle_packet(struct nf_conn *ct,
return nf_conntrack_dccp_packet(ct, skb, dataoff,
ctinfo, state);
#endif
+#ifdef CONFIG_NF_CT_PROTO_GRE
+ case IPPROTO_GRE:
+ return nf_conntrack_gre_packet(ct, skb, dataoff,
+ ctinfo, state);
+#endif
}
- WARN_ON_ONCE(1);
- return -NF_ACCEPT;
+ return generic_packet(ct, skb, ctinfo);
}
unsigned int
@@ -1634,11 +1650,7 @@ nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
goto out;
}
- if (l4proto->packet)
- ret = l4proto->packet(ct, skb, dataoff, ctinfo, state);
- else
- ret = nf_conntrack_handle_packet(ct, skb, dataoff, ctinfo, state);
-
+ ret = nf_conntrack_handle_packet(ct, skb, dataoff, ctinfo, state);
if (ret <= 0) {
/* Invalid: inverse of the return code tells
* the netfilter core what to do */
diff --git a/net/netfilter/nf_conntrack_proto_generic.c b/net/netfilter/nf_conntrack_proto_generic.c
index 5a5bf7cb6508..78358d192c60 100644
--- a/net/netfilter/nf_conntrack_proto_generic.c
+++ b/net/netfilter/nf_conntrack_proto_generic.c
@@ -15,40 +15,6 @@
static const unsigned int nf_ct_generic_timeout = 600*HZ;
-static bool nf_generic_should_process(u8 proto)
-{
- switch (proto) {
-#ifdef CONFIG_NF_CT_PROTO_GRE_MODULE
- case IPPROTO_GRE:
- return false;
-#endif
- default:
- return true;
- }
-}
-
-/* Returns verdict for packet, or -1 for invalid. */
-static int generic_packet(struct nf_conn *ct,
- struct sk_buff *skb,
- unsigned int dataoff,
- enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state)
-{
- const unsigned int *timeout = nf_ct_timeout_lookup(ct);
-
- if (!nf_generic_should_process(nf_ct_protonum(ct))) {
- pr_warn_once("conntrack: generic helper won't handle protocol %d. Please consider loading the specific helper module.\n",
- nf_ct_protonum(ct));
- return -NF_ACCEPT;
- }
-
- if (!timeout)
- timeout = &nf_generic_pernet(nf_ct_net(ct))->timeout;
-
- nf_ct_refresh_acct(ct, ctinfo, skb, *timeout);
- return NF_ACCEPT;
-}
-
#ifdef CONFIG_NF_CONNTRACK_TIMEOUT
#include <linux/netfilter/nfnetlink.h>
@@ -139,7 +105,6 @@ static struct nf_proto_net *generic_get_net_proto(struct net *net)
const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic =
{
.l4proto = 255,
- .packet = generic_packet,
#ifdef CONFIG_NF_CONNTRACK_TIMEOUT
.ctnl_timeout = {
.nlattr_to_obj = generic_timeout_nlattr_to_obj,
diff --git a/net/netfilter/nf_conntrack_proto_gre.c b/net/netfilter/nf_conntrack_proto_gre.c
index 0c26cfd5d29f..f6391991bcf6 100644
--- a/net/netfilter/nf_conntrack_proto_gre.c
+++ b/net/netfilter/nf_conntrack_proto_gre.c
@@ -213,11 +213,11 @@ static unsigned int *gre_get_timeouts(struct net *net)
}
/* Returns verdict for packet, and may modify conntrack */
-static int gre_packet(struct nf_conn *ct,
- struct sk_buff *skb,
- unsigned int dataoff,
- enum ip_conntrack_info ctinfo,
- const struct nf_hook_state *state)
+int nf_conntrack_gre_packet(struct nf_conn *ct,
+ struct sk_buff *skb,
+ unsigned int dataoff,
+ enum ip_conntrack_info ctinfo,
+ const struct nf_hook_state *state)
{
if (state->pf != NFPROTO_IPV4)
return -NF_ACCEPT;
@@ -371,7 +371,6 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_gre = {
#ifdef CONFIG_NF_CONNTRACK_PROCFS
.print_conntrack = gre_print_conntrack,
#endif
- .packet = gre_packet,
.destroy = gre_destroy,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
.tuple_to_nlattr = nf_ct_port_tuple_to_nlattr,
--
2.11.0
^ permalink raw reply related
* [PATCH 19/33] netfilter: conntrack: remove sysctl registration helpers
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
After previous patch these are not used anymore.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netns/conntrack.h | 4 --
net/netfilter/nf_conntrack_proto.c | 77 +-------------------------------------
2 files changed, 1 insertion(+), 80 deletions(-)
diff --git a/include/net/netns/conntrack.h b/include/net/netns/conntrack.h
index c72f413a2d4d..b8faa72e0f5b 100644
--- a/include/net/netns/conntrack.h
+++ b/include/net/netns/conntrack.h
@@ -19,10 +19,6 @@ struct ctl_table_header;
struct nf_conntrack_ecache;
struct nf_proto_net {
-#ifdef CONFIG_SYSCTL
- struct ctl_table_header *ctl_table_header;
- struct ctl_table *ctl_table;
-#endif
unsigned int users;
};
diff --git a/net/netfilter/nf_conntrack_proto.c b/net/netfilter/nf_conntrack_proto.c
index 13b782cc6a7a..c8f024d731c8 100644
--- a/net/netfilter/nf_conntrack_proto.c
+++ b/net/netfilter/nf_conntrack_proto.c
@@ -48,35 +48,6 @@ static struct nf_conntrack_l4proto __rcu *nf_ct_protos[MAX_NF_CT_PROTO + 1] __re
static DEFINE_MUTEX(nf_ct_proto_mutex);
#ifdef CONFIG_SYSCTL
-static int
-nf_ct_register_sysctl(struct net *net,
- struct ctl_table_header **header,
- const char *path,
- struct ctl_table *table)
-{
- if (*header == NULL) {
- *header = register_net_sysctl(net, path, table);
- if (*header == NULL)
- return -ENOMEM;
- }
-
- return 0;
-}
-
-static void
-nf_ct_unregister_sysctl(struct ctl_table_header **header,
- struct ctl_table **table,
- unsigned int users)
-{
- if (users > 0)
- return;
-
- unregister_net_sysctl_table(*header);
- kfree(*table);
- *header = NULL;
- *table = NULL;
-}
-
__printf(5, 6)
void nf_l4proto_log_invalid(const struct sk_buff *skb,
struct net *net,
@@ -161,40 +132,6 @@ static struct nf_proto_net *nf_ct_l4proto_net(struct net *net,
return NULL;
}
-static
-int nf_ct_l4proto_register_sysctl(struct net *net,
- struct nf_proto_net *pn)
-{
- int err = 0;
-
-#ifdef CONFIG_SYSCTL
- if (pn->ctl_table != NULL) {
- err = nf_ct_register_sysctl(net,
- &pn->ctl_table_header,
- "net/netfilter",
- pn->ctl_table);
- if (err < 0) {
- if (!pn->users) {
- kfree(pn->ctl_table);
- pn->ctl_table = NULL;
- }
- }
- }
-#endif /* CONFIG_SYSCTL */
- return err;
-}
-
-static
-void nf_ct_l4proto_unregister_sysctl(struct nf_proto_net *pn)
-{
-#ifdef CONFIG_SYSCTL
- if (pn->ctl_table_header != NULL)
- nf_ct_unregister_sysctl(&pn->ctl_table_header,
- &pn->ctl_table,
- pn->users);
-#endif /* CONFIG_SYSCTL */
-}
-
/* FIXME: Allow NULL functions and sub in pointers to generic for
them. --RR */
int nf_ct_l4proto_register_one(const struct nf_conntrack_l4proto *l4proto)
@@ -237,10 +174,6 @@ static int nf_ct_l4proto_pernet_register_one(struct net *net,
if (pn == NULL)
goto out;
- ret = nf_ct_l4proto_register_sysctl(net, pn);
- if (ret < 0)
- goto out;
-
pn->users++;
out:
return ret;
@@ -280,7 +213,6 @@ static void nf_ct_l4proto_pernet_unregister_one(struct net *net,
return;
pn->users--;
- nf_ct_l4proto_unregister_sysctl(pn);
}
static void
@@ -859,17 +791,11 @@ int nf_conntrack_proto_pernet_init(struct net *net)
err = nf_conntrack_l4proto_generic.init_net(net);
if (err < 0)
return err;
- err = nf_ct_l4proto_register_sysctl(net,
- pn);
- if (err < 0)
- return err;
err = nf_ct_l4proto_pernet_register(net, builtin_l4proto,
ARRAY_SIZE(builtin_l4proto));
- if (err < 0) {
- nf_ct_l4proto_unregister_sysctl(pn);
+ if (err < 0)
return err;
- }
pn->users++;
return 0;
@@ -883,7 +809,6 @@ void nf_conntrack_proto_pernet_fini(struct net *net)
nf_ct_l4proto_pernet_unregister(net, builtin_l4proto,
ARRAY_SIZE(builtin_l4proto));
pn->users--;
- nf_ct_l4proto_unregister_sysctl(pn);
#ifdef CONFIG_NF_CT_PROTO_GRE
nf_ct_gre_keymap_flush(net);
#endif
--
2.11.0
^ permalink raw reply related
* [PATCH 20/33] netfilter: conntrack: remove l4proto init and get_net callbacks
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
Those were needed we still had modular trackers.
As we don't have those anymore, prefer direct calls and remove all
the (un)register infrastructure associated with this.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_conntrack_l4proto.h | 15 ++--
include/net/netns/conntrack.h | 11 ---
net/netfilter/nf_conntrack_proto.c | 103 ++++-----------------------
net/netfilter/nf_conntrack_proto_dccp.c | 42 ++++-------
net/netfilter/nf_conntrack_proto_generic.c | 11 +--
net/netfilter/nf_conntrack_proto_gre.c | 5 +-
net/netfilter/nf_conntrack_proto_icmp.c | 12 +---
net/netfilter/nf_conntrack_proto_icmpv6.c | 12 +---
net/netfilter/nf_conntrack_proto_sctp.c | 29 +++-----
net/netfilter/nf_conntrack_proto_tcp.c | 35 +++------
net/netfilter/nf_conntrack_proto_udp.c | 24 ++-----
11 files changed, 65 insertions(+), 234 deletions(-)
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index 87d9c198c05e..d5909e51ca92 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -61,12 +61,6 @@ struct nf_conntrack_l4proto {
/* Print out the private part of the conntrack. */
void (*print_conntrack)(struct seq_file *s, struct nf_conn *);
#endif
-
- /* Init l4proto pernet data */
- int (*init_net)(struct net *net);
-
- /* Return the per-net protocol part. */
- struct nf_proto_net *(*get_net_proto)(struct net *net);
};
bool icmp_pkt_to_tuple(const struct sk_buff *skb,
@@ -135,6 +129,15 @@ int nf_conntrack_gre_packet(struct nf_conn *ct,
enum ip_conntrack_info ctinfo,
const struct nf_hook_state *state);
+void nf_conntrack_generic_init_net(struct net *net);
+void nf_conntrack_tcp_init_net(struct net *net);
+void nf_conntrack_udp_init_net(struct net *net);
+void nf_conntrack_gre_init_net(struct net *net);
+void nf_conntrack_dccp_init_net(struct net *net);
+void nf_conntrack_sctp_init_net(struct net *net);
+void nf_conntrack_icmp_init_net(struct net *net);
+void nf_conntrack_icmpv6_init_net(struct net *net);
+
/* Existing built-in generic protocol */
extern const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic;
diff --git a/include/net/netns/conntrack.h b/include/net/netns/conntrack.h
index b8faa72e0f5b..f19b53130bf7 100644
--- a/include/net/netns/conntrack.h
+++ b/include/net/netns/conntrack.h
@@ -18,17 +18,11 @@
struct ctl_table_header;
struct nf_conntrack_ecache;
-struct nf_proto_net {
- unsigned int users;
-};
-
struct nf_generic_net {
- struct nf_proto_net pn;
unsigned int timeout;
};
struct nf_tcp_net {
- struct nf_proto_net pn;
unsigned int timeouts[TCP_CONNTRACK_TIMEOUT_MAX];
unsigned int tcp_loose;
unsigned int tcp_be_liberal;
@@ -42,18 +36,15 @@ enum udp_conntrack {
};
struct nf_udp_net {
- struct nf_proto_net pn;
unsigned int timeouts[UDP_CT_MAX];
};
struct nf_icmp_net {
- struct nf_proto_net pn;
unsigned int timeout;
};
#ifdef CONFIG_NF_CT_PROTO_DCCP
struct nf_dccp_net {
- struct nf_proto_net pn;
int dccp_loose;
unsigned int dccp_timeout[CT_DCCP_MAX + 1];
};
@@ -61,7 +52,6 @@ struct nf_dccp_net {
#ifdef CONFIG_NF_CT_PROTO_SCTP
struct nf_sctp_net {
- struct nf_proto_net pn;
unsigned int timeouts[SCTP_CONNTRACK_MAX];
};
#endif
@@ -74,7 +64,6 @@ enum gre_conntrack {
};
struct nf_gre_net {
- struct nf_proto_net nf;
struct list_head keymap_list;
unsigned int timeouts[GRE_CT_MAX];
};
diff --git a/net/netfilter/nf_conntrack_proto.c b/net/netfilter/nf_conntrack_proto.c
index c8f024d731c8..e6bc02c13f0f 100644
--- a/net/netfilter/nf_conntrack_proto.c
+++ b/net/netfilter/nf_conntrack_proto.c
@@ -123,15 +123,6 @@ static int kill_l4proto(struct nf_conn *i, void *data)
return nf_ct_protonum(i) == l4proto->l4proto;
}
-static struct nf_proto_net *nf_ct_l4proto_net(struct net *net,
- const struct nf_conntrack_l4proto *l4proto)
-{
- if (l4proto->get_net_proto)
- return l4proto->get_net_proto(net);
-
- return NULL;
-}
-
/* FIXME: Allow NULL functions and sub in pointers to generic for
them. --RR */
int nf_ct_l4proto_register_one(const struct nf_conntrack_l4proto *l4proto)
@@ -158,27 +149,6 @@ int nf_ct_l4proto_register_one(const struct nf_conntrack_l4proto *l4proto)
}
EXPORT_SYMBOL_GPL(nf_ct_l4proto_register_one);
-static int nf_ct_l4proto_pernet_register_one(struct net *net,
- const struct nf_conntrack_l4proto *l4proto)
-{
- int ret = 0;
- struct nf_proto_net *pn = NULL;
-
- if (l4proto->init_net) {
- ret = l4proto->init_net(net);
- if (ret < 0)
- goto out;
- }
-
- pn = nf_ct_l4proto_net(net, l4proto);
- if (pn == NULL)
- goto out;
-
- pn->users++;
-out:
- return ret;
-}
-
static void __nf_ct_l4proto_unregister_one(const struct nf_conntrack_l4proto *l4proto)
{
@@ -204,17 +174,6 @@ void nf_ct_l4proto_unregister_one(const struct nf_conntrack_l4proto *l4proto)
}
EXPORT_SYMBOL_GPL(nf_ct_l4proto_unregister_one);
-static void nf_ct_l4proto_pernet_unregister_one(struct net *net,
- const struct nf_conntrack_l4proto *l4proto)
-{
- struct nf_proto_net *pn = nf_ct_l4proto_net(net, l4proto);
-
- if (pn == NULL)
- return;
-
- pn->users--;
-}
-
static void
nf_ct_l4proto_unregister(const struct nf_conntrack_l4proto * const l4proto[],
unsigned int num_proto)
@@ -252,34 +211,6 @@ nf_ct_l4proto_register(const struct nf_conntrack_l4proto * const l4proto[],
return ret;
}
-static void nf_ct_l4proto_pernet_unregister(struct net *net,
- const struct nf_conntrack_l4proto *const l4proto[],
- unsigned int num_proto)
-{
- while (num_proto-- != 0)
- nf_ct_l4proto_pernet_unregister_one(net, l4proto[num_proto]);
-}
-
-static int nf_ct_l4proto_pernet_register(struct net *net,
- const struct nf_conntrack_l4proto *const l4proto[],
- unsigned int num_proto)
-{
- int ret = -EINVAL;
- unsigned int i;
-
- for (i = 0; i < num_proto; i++) {
- ret = nf_ct_l4proto_pernet_register_one(net, l4proto[i]);
- if (ret < 0)
- break;
- }
- if (i != num_proto) {
- pr_err("nf_conntrack %d: pernet registration failed\n",
- l4proto[i]->l4proto);
- nf_ct_l4proto_pernet_unregister(net, l4proto, i);
- }
- return ret;
-}
-
static unsigned int nf_confirm(struct sk_buff *skb,
unsigned int protoff,
struct nf_conn *ct,
@@ -784,31 +715,25 @@ void nf_conntrack_proto_fini(void)
int nf_conntrack_proto_pernet_init(struct net *net)
{
- int err;
- struct nf_proto_net *pn = nf_ct_l4proto_net(net,
- &nf_conntrack_l4proto_generic);
-
- err = nf_conntrack_l4proto_generic.init_net(net);
- if (err < 0)
- return err;
-
- err = nf_ct_l4proto_pernet_register(net, builtin_l4proto,
- ARRAY_SIZE(builtin_l4proto));
- if (err < 0)
- return err;
-
- pn->users++;
+ nf_conntrack_generic_init_net(net);
+ nf_conntrack_udp_init_net(net);
+ nf_conntrack_tcp_init_net(net);
+ nf_conntrack_icmp_init_net(net);
+ nf_conntrack_icmpv6_init_net(net);
+#ifdef CONFIG_NF_CT_PROTO_DCCP
+ nf_conntrack_dccp_init_net(net);
+#endif
+#ifdef CONFIG_NF_CT_PROTO_SCTP
+ nf_conntrack_sctp_init_net(net);
+#endif
+#ifdef CONFIG_NF_CT_PROTO_GRE
+ nf_conntrack_gre_init_net(net);
+#endif
return 0;
}
void nf_conntrack_proto_pernet_fini(struct net *net)
{
- struct nf_proto_net *pn = nf_ct_l4proto_net(net,
- &nf_conntrack_l4proto_generic);
-
- nf_ct_l4proto_pernet_unregister(net, builtin_l4proto,
- ARRAY_SIZE(builtin_l4proto));
- pn->users--;
#ifdef CONFIG_NF_CT_PROTO_GRE
nf_ct_gre_keymap_flush(net);
#endif
diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c
index 31ba88311bc4..6fca80587505 100644
--- a/net/netfilter/nf_conntrack_proto_dccp.c
+++ b/net/netfilter/nf_conntrack_proto_dccp.c
@@ -724,34 +724,24 @@ dccp_timeout_nla_policy[CTA_TIMEOUT_DCCP_MAX+1] = {
};
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
-static int dccp_init_net(struct net *net)
+void nf_conntrack_dccp_init_net(struct net *net)
{
struct nf_dccp_net *dn = nf_dccp_pernet(net);
- struct nf_proto_net *pn = &dn->pn;
-
- if (!pn->users) {
- /* default values */
- dn->dccp_loose = 1;
- dn->dccp_timeout[CT_DCCP_REQUEST] = 2 * DCCP_MSL;
- dn->dccp_timeout[CT_DCCP_RESPOND] = 4 * DCCP_MSL;
- dn->dccp_timeout[CT_DCCP_PARTOPEN] = 4 * DCCP_MSL;
- dn->dccp_timeout[CT_DCCP_OPEN] = 12 * 3600 * HZ;
- dn->dccp_timeout[CT_DCCP_CLOSEREQ] = 64 * HZ;
- dn->dccp_timeout[CT_DCCP_CLOSING] = 64 * HZ;
- dn->dccp_timeout[CT_DCCP_TIMEWAIT] = 2 * DCCP_MSL;
-
- /* timeouts[0] is unused, make it same as SYN_SENT so
- * ->timeouts[0] contains 'new' timeout, like udp or icmp.
- */
- dn->dccp_timeout[CT_DCCP_NONE] = dn->dccp_timeout[CT_DCCP_REQUEST];
- }
-
- return 0;
-}
-static struct nf_proto_net *dccp_get_net_proto(struct net *net)
-{
- return &net->ct.nf_ct_proto.dccp.pn;
+ /* default values */
+ dn->dccp_loose = 1;
+ dn->dccp_timeout[CT_DCCP_REQUEST] = 2 * DCCP_MSL;
+ dn->dccp_timeout[CT_DCCP_RESPOND] = 4 * DCCP_MSL;
+ dn->dccp_timeout[CT_DCCP_PARTOPEN] = 4 * DCCP_MSL;
+ dn->dccp_timeout[CT_DCCP_OPEN] = 12 * 3600 * HZ;
+ dn->dccp_timeout[CT_DCCP_CLOSEREQ] = 64 * HZ;
+ dn->dccp_timeout[CT_DCCP_CLOSING] = 64 * HZ;
+ dn->dccp_timeout[CT_DCCP_TIMEWAIT] = 2 * DCCP_MSL;
+
+ /* timeouts[0] is unused, make it same as SYN_SENT so
+ * ->timeouts[0] contains 'new' timeout, like udp or icmp.
+ */
+ dn->dccp_timeout[CT_DCCP_NONE] = dn->dccp_timeout[CT_DCCP_REQUEST];
}
const struct nf_conntrack_l4proto nf_conntrack_l4proto_dccp = {
@@ -778,6 +768,4 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_dccp = {
.nla_policy = dccp_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .init_net = dccp_init_net,
- .get_net_proto = dccp_get_net_proto,
};
diff --git a/net/netfilter/nf_conntrack_proto_generic.c b/net/netfilter/nf_conntrack_proto_generic.c
index 0edbf82594d0..0f526fafecae 100644
--- a/net/netfilter/nf_conntrack_proto_generic.c
+++ b/net/netfilter/nf_conntrack_proto_generic.c
@@ -60,18 +60,11 @@ generic_timeout_nla_policy[CTA_TIMEOUT_GENERIC_MAX+1] = {
};
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
-static int generic_init_net(struct net *net)
+void nf_conntrack_generic_init_net(struct net *net)
{
struct nf_generic_net *gn = nf_generic_pernet(net);
gn->timeout = nf_ct_generic_timeout;
-
- return 0;
-}
-
-static struct nf_proto_net *generic_get_net_proto(struct net *net)
-{
- return &net->ct.nf_ct_proto.generic.pn;
}
const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic =
@@ -86,6 +79,4 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic =
.nla_policy = generic_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .init_net = generic_init_net,
- .get_net_proto = generic_get_net_proto,
};
diff --git a/net/netfilter/nf_conntrack_proto_gre.c b/net/netfilter/nf_conntrack_proto_gre.c
index e573ec0fa12b..fa765d2fd586 100644
--- a/net/netfilter/nf_conntrack_proto_gre.c
+++ b/net/netfilter/nf_conntrack_proto_gre.c
@@ -313,7 +313,7 @@ gre_timeout_nla_policy[CTA_TIMEOUT_GRE_MAX+1] = {
};
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
-static int gre_init_net(struct net *net)
+void nf_conntrack_gre_init_net(struct net *net)
{
struct nf_gre_net *net_gre = gre_pernet(net);
int i;
@@ -321,8 +321,6 @@ static int gre_init_net(struct net *net)
INIT_LIST_HEAD(&net_gre->keymap_list);
for (i = 0; i < GRE_CT_MAX; i++)
net_gre->timeouts[i] = gre_timeouts[i];
-
- return 0;
}
/* protocol helper struct */
@@ -347,5 +345,4 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_gre = {
.nla_policy = gre_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .init_net = gre_init_net,
};
diff --git a/net/netfilter/nf_conntrack_proto_icmp.c b/net/netfilter/nf_conntrack_proto_icmp.c
index eb77f747759f..7df477996b16 100644
--- a/net/netfilter/nf_conntrack_proto_icmp.c
+++ b/net/netfilter/nf_conntrack_proto_icmp.c
@@ -298,19 +298,11 @@ icmp_timeout_nla_policy[CTA_TIMEOUT_ICMP_MAX+1] = {
};
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
-
-static int icmp_init_net(struct net *net)
+void nf_conntrack_icmp_init_net(struct net *net)
{
struct nf_icmp_net *in = nf_icmp_pernet(net);
in->timeout = nf_ct_icmp_timeout;
-
- return 0;
-}
-
-static struct nf_proto_net *icmp_get_net_proto(struct net *net)
-{
- return &net->ct.nf_ct_proto.icmp.pn;
}
const struct nf_conntrack_l4proto nf_conntrack_l4proto_icmp =
@@ -331,6 +323,4 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_icmp =
.nla_policy = icmp_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .init_net = icmp_init_net,
- .get_net_proto = icmp_get_net_proto,
};
diff --git a/net/netfilter/nf_conntrack_proto_icmpv6.c b/net/netfilter/nf_conntrack_proto_icmpv6.c
index d243ef8a128e..bec4a3211658 100644
--- a/net/netfilter/nf_conntrack_proto_icmpv6.c
+++ b/net/netfilter/nf_conntrack_proto_icmpv6.c
@@ -309,19 +309,11 @@ icmpv6_timeout_nla_policy[CTA_TIMEOUT_ICMPV6_MAX+1] = {
};
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
-
-static int icmpv6_init_net(struct net *net)
+void nf_conntrack_icmpv6_init_net(struct net *net)
{
struct nf_icmp_net *in = nf_icmpv6_pernet(net);
in->timeout = nf_ct_icmpv6_timeout;
-
- return 0;
-}
-
-static struct nf_proto_net *icmpv6_get_net_proto(struct net *net)
-{
- return &net->ct.nf_ct_proto.icmpv6.pn;
}
const struct nf_conntrack_l4proto nf_conntrack_l4proto_icmpv6 =
@@ -342,6 +334,4 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_icmpv6 =
.nla_policy = icmpv6_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .init_net = icmpv6_init_net,
- .get_net_proto = icmpv6_get_net_proto,
};
diff --git a/net/netfilter/nf_conntrack_proto_sctp.c b/net/netfilter/nf_conntrack_proto_sctp.c
index 31130f218d8f..a7818101ad80 100644
--- a/net/netfilter/nf_conntrack_proto_sctp.c
+++ b/net/netfilter/nf_conntrack_proto_sctp.c
@@ -642,29 +642,18 @@ sctp_timeout_nla_policy[CTA_TIMEOUT_SCTP_MAX+1] = {
};
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
-static int sctp_init_net(struct net *net)
+void nf_conntrack_sctp_init_net(struct net *net)
{
struct nf_sctp_net *sn = nf_sctp_pernet(net);
- struct nf_proto_net *pn = &sn->pn;
-
- if (!pn->users) {
- int i;
-
- for (i = 0; i < SCTP_CONNTRACK_MAX; i++)
- sn->timeouts[i] = sctp_timeouts[i];
-
- /* timeouts[0] is unused, init it so ->timeouts[0] contains
- * 'new' timeout, like udp or icmp.
- */
- sn->timeouts[0] = sctp_timeouts[SCTP_CONNTRACK_CLOSED];
- }
+ int i;
- return 0;
-}
+ for (i = 0; i < SCTP_CONNTRACK_MAX; i++)
+ sn->timeouts[i] = sctp_timeouts[i];
-static struct nf_proto_net *sctp_get_net_proto(struct net *net)
-{
- return &net->ct.nf_ct_proto.sctp.pn;
+ /* timeouts[0] is unused, init it so ->timeouts[0] contains
+ * 'new' timeout, like udp or icmp.
+ */
+ sn->timeouts[0] = sctp_timeouts[SCTP_CONNTRACK_CLOSED];
}
const struct nf_conntrack_l4proto nf_conntrack_l4proto_sctp = {
@@ -691,6 +680,4 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_sctp = {
.nla_policy = sctp_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .init_net = sctp_init_net,
- .get_net_proto = sctp_get_net_proto,
};
diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c
index 422bdedac0ed..01c748fa8913 100644
--- a/net/netfilter/nf_conntrack_proto_tcp.c
+++ b/net/netfilter/nf_conntrack_proto_tcp.c
@@ -1387,32 +1387,21 @@ static const struct nla_policy tcp_timeout_nla_policy[CTA_TIMEOUT_TCP_MAX+1] = {
};
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
-static int tcp_init_net(struct net *net)
+void nf_conntrack_tcp_init_net(struct net *net)
{
struct nf_tcp_net *tn = nf_tcp_pernet(net);
- struct nf_proto_net *pn = &tn->pn;
-
- if (!pn->users) {
- int i;
-
- for (i = 0; i < TCP_CONNTRACK_TIMEOUT_MAX; i++)
- tn->timeouts[i] = tcp_timeouts[i];
-
- /* timeouts[0] is unused, make it same as SYN_SENT so
- * ->timeouts[0] contains 'new' timeout, like udp or icmp.
- */
- tn->timeouts[0] = tcp_timeouts[TCP_CONNTRACK_SYN_SENT];
- tn->tcp_loose = nf_ct_tcp_loose;
- tn->tcp_be_liberal = nf_ct_tcp_be_liberal;
- tn->tcp_max_retrans = nf_ct_tcp_max_retrans;
- }
+ int i;
- return 0;
-}
+ for (i = 0; i < TCP_CONNTRACK_TIMEOUT_MAX; i++)
+ tn->timeouts[i] = tcp_timeouts[i];
-static struct nf_proto_net *tcp_get_net_proto(struct net *net)
-{
- return &net->ct.nf_ct_proto.tcp.pn;
+ /* timeouts[0] is unused, make it same as SYN_SENT so
+ * ->timeouts[0] contains 'new' timeout, like udp or icmp.
+ */
+ tn->timeouts[0] = tcp_timeouts[TCP_CONNTRACK_SYN_SENT];
+ tn->tcp_loose = nf_ct_tcp_loose;
+ tn->tcp_be_liberal = nf_ct_tcp_be_liberal;
+ tn->tcp_max_retrans = nf_ct_tcp_max_retrans;
}
const struct nf_conntrack_l4proto nf_conntrack_l4proto_tcp =
@@ -1441,6 +1430,4 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_tcp =
.nla_policy = tcp_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .init_net = tcp_init_net,
- .get_net_proto = tcp_get_net_proto,
};
diff --git a/net/netfilter/nf_conntrack_proto_udp.c b/net/netfilter/nf_conntrack_proto_udp.c
index 6e81e79844d7..951366dfbec3 100644
--- a/net/netfilter/nf_conntrack_proto_udp.c
+++ b/net/netfilter/nf_conntrack_proto_udp.c
@@ -260,25 +260,13 @@ udp_timeout_nla_policy[CTA_TIMEOUT_UDP_MAX+1] = {
};
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
-
-static int udp_init_net(struct net *net)
+void nf_conntrack_udp_init_net(struct net *net)
{
struct nf_udp_net *un = nf_udp_pernet(net);
- struct nf_proto_net *pn = &un->pn;
-
- if (!pn->users) {
- int i;
-
- for (i = 0; i < UDP_CT_MAX; i++)
- un->timeouts[i] = udp_timeouts[i];
- }
+ int i;
- return 0;
-}
-
-static struct nf_proto_net *udp_get_net_proto(struct net *net)
-{
- return &net->ct.nf_ct_proto.udp.pn;
+ for (i = 0; i < UDP_CT_MAX; i++)
+ un->timeouts[i] = udp_timeouts[i];
}
const struct nf_conntrack_l4proto nf_conntrack_l4proto_udp =
@@ -300,8 +288,6 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_udp =
.nla_policy = udp_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .init_net = udp_init_net,
- .get_net_proto = udp_get_net_proto,
};
#ifdef CONFIG_NF_CT_PROTO_UDPLITE
@@ -324,7 +310,5 @@ const struct nf_conntrack_l4proto nf_conntrack_l4proto_udplite =
.nla_policy = udp_timeout_nla_policy,
},
#endif /* CONFIG_NF_CONNTRACK_TIMEOUT */
- .init_net = udp_init_net,
- .get_net_proto = udp_get_net_proto,
};
#endif
--
2.11.0
^ permalink raw reply related
* [PATCH 23/33] netfilter: nf_conntrack: provide modparam to always register conntrack hooks
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
The connection tracking hooks can be optionally registered per netns
when conntrack is specifically invoked from the ruleset since
0c66dc1ea3f0 ("netfilter: conntrack: register hooks in netns when needed
by ruleset"). Then, since 4d3a57f23dec ("netfilter: conntrack: do not
enable connection tracking unless needed"), the default behaviour is
changed to always register them on demand.
This patch provides a toggle that allows users to always register them.
Without this toggle, in order to use conntrack for statistics
collection, you need a dummy rule that refers to conntrack, eg.
iptables -I INPUT -m state --state NEW
This patch allows users to restore the original behaviour via modparam,
ie. always register connection tracking, eg.
modprobe nf_conntrack enable_hooks=1
Hence, no dummy rule is required.
Reported-by: Laura Garcia <nevola@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
net/netfilter/nf_conntrack_standalone.c | 28 ++++++++++++++++++++++++----
1 file changed, 24 insertions(+), 4 deletions(-)
diff --git a/net/netfilter/nf_conntrack_standalone.c b/net/netfilter/nf_conntrack_standalone.c
index ddfca5f1784c..8928a4d0933e 100644
--- a/net/netfilter/nf_conntrack_standalone.c
+++ b/net/netfilter/nf_conntrack_standalone.c
@@ -24,6 +24,10 @@
#include <net/netfilter/nf_conntrack_timestamp.h>
#include <linux/rculist_nulls.h>
+static bool enable_hooks __read_mostly;
+MODULE_PARM_DESC(enable_hooks, "Always enable conntrack hooks");
+module_param(enable_hooks, bool, 0000);
+
unsigned int nf_conntrack_net_id __read_mostly;
#ifdef CONFIG_NF_CONNTRACK_PROCFS
@@ -1075,6 +1079,15 @@ static void nf_conntrack_standalone_fini_sysctl(struct net *net)
}
#endif /* CONFIG_SYSCTL */
+static void nf_conntrack_fini_net(struct net *net)
+{
+ if (enable_hooks)
+ nf_ct_netns_put(net, NFPROTO_INET);
+
+ nf_conntrack_standalone_fini_proc(net);
+ nf_conntrack_standalone_fini_sysctl(net);
+}
+
static int nf_conntrack_pernet_init(struct net *net)
{
int ret;
@@ -1093,8 +1106,16 @@ static int nf_conntrack_pernet_init(struct net *net)
if (ret < 0)
goto out_init_net;
+ if (enable_hooks) {
+ ret = nf_ct_netns_get(net, NFPROTO_INET);
+ if (ret < 0)
+ goto out_hooks;
+ }
+
return 0;
+out_hooks:
+ nf_conntrack_fini_net(net);
out_init_net:
nf_conntrack_standalone_fini_proc(net);
out_proc:
@@ -1106,10 +1127,9 @@ static void nf_conntrack_pernet_exit(struct list_head *net_exit_list)
{
struct net *net;
- list_for_each_entry(net, net_exit_list, exit_list) {
- nf_conntrack_standalone_fini_sysctl(net);
- nf_conntrack_standalone_fini_proc(net);
- }
+ list_for_each_entry(net, net_exit_list, exit_list)
+ nf_conntrack_fini_net(net);
+
nf_conntrack_cleanup_net_list(net_exit_list);
}
--
2.11.0
^ permalink raw reply related
* [PATCH 17/33] netfilter: conntrack: avoid unneeded nf_conntrack_l4proto lookups
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
after removal of the packet and invert function pointers, several
places do not need to lookup the l4proto structure anymore.
Remove those lookups.
The function nf_ct_invert_tuplepr becomes redundant, replace
it with nf_ct_invert_tuple everywhere.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_conntrack.h | 2 --
include/net/netfilter/nf_conntrack_core.h | 3 +-
net/ipv4/netfilter/nf_nat_l3proto_ipv4.c | 2 +-
net/ipv6/netfilter/nf_nat_l3proto_ipv6.c | 2 +-
net/netfilter/nf_conntrack_core.c | 53 ++++++-------------------------
net/netfilter/nf_conntrack_pptp.c | 2 +-
net/netfilter/nf_conntrack_proto_icmp.c | 6 +---
net/netfilter/nf_conntrack_proto_icmpv6.c | 6 +---
net/netfilter/nf_nat_core.c | 12 +++----
net/openvswitch/conntrack.c | 2 +-
10 files changed, 22 insertions(+), 68 deletions(-)
diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h
index 249d0a5b12b8..b5aac5ae5129 100644
--- a/include/net/netfilter/nf_conntrack.h
+++ b/include/net/netfilter/nf_conntrack.h
@@ -187,8 +187,6 @@ bool nf_ct_delete(struct nf_conn *ct, u32 pid, int report);
bool nf_ct_get_tuplepr(const struct sk_buff *skb, unsigned int nhoff,
u_int16_t l3num, struct net *net,
struct nf_conntrack_tuple *tuple);
-bool nf_ct_invert_tuplepr(struct nf_conntrack_tuple *inverse,
- const struct nf_conntrack_tuple *orig);
void __nf_ct_refresh_acct(struct nf_conn *ct, enum ip_conntrack_info ctinfo,
const struct sk_buff *skb,
diff --git a/include/net/netfilter/nf_conntrack_core.h b/include/net/netfilter/nf_conntrack_core.h
index afc9b3620473..235c182022b2 100644
--- a/include/net/netfilter/nf_conntrack_core.h
+++ b/include/net/netfilter/nf_conntrack_core.h
@@ -39,8 +39,7 @@ void nf_conntrack_init_end(void);
void nf_conntrack_cleanup_end(void);
bool nf_ct_invert_tuple(struct nf_conntrack_tuple *inverse,
- const struct nf_conntrack_tuple *orig,
- const struct nf_conntrack_l4proto *l4proto);
+ const struct nf_conntrack_tuple *orig);
/* Find a connection corresponding to a tuple. */
struct nf_conntrack_tuple_hash *
diff --git a/net/ipv4/netfilter/nf_nat_l3proto_ipv4.c b/net/ipv4/netfilter/nf_nat_l3proto_ipv4.c
index 2687db015b6f..e26165af45cb 100644
--- a/net/ipv4/netfilter/nf_nat_l3proto_ipv4.c
+++ b/net/ipv4/netfilter/nf_nat_l3proto_ipv4.c
@@ -214,7 +214,7 @@ int nf_nat_icmp_reply_translation(struct sk_buff *skb,
}
/* Change outer to look like the reply to an incoming packet */
- nf_ct_invert_tuplepr(&target, &ct->tuplehash[!dir].tuple);
+ nf_ct_invert_tuple(&target, &ct->tuplehash[!dir].tuple);
if (!nf_nat_ipv4_manip_pkt(skb, 0, &target, manip))
return 0;
diff --git a/net/ipv6/netfilter/nf_nat_l3proto_ipv6.c b/net/ipv6/netfilter/nf_nat_l3proto_ipv6.c
index 23022447eb49..9c914db44bec 100644
--- a/net/ipv6/netfilter/nf_nat_l3proto_ipv6.c
+++ b/net/ipv6/netfilter/nf_nat_l3proto_ipv6.c
@@ -225,7 +225,7 @@ int nf_nat_icmpv6_reply_translation(struct sk_buff *skb,
skb->len - hdrlen, 0));
}
- nf_ct_invert_tuplepr(&target, &ct->tuplehash[!dir].tuple);
+ nf_ct_invert_tuple(&target, &ct->tuplehash[!dir].tuple);
if (!nf_nat_ipv6_manip_pkt(skb, 0, &target, manip))
return 0;
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 264074f04615..728d2b5bdb1a 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -229,8 +229,7 @@ nf_ct_get_tuple(const struct sk_buff *skb,
u_int16_t l3num,
u_int8_t protonum,
struct net *net,
- struct nf_conntrack_tuple *tuple,
- const struct nf_conntrack_l4proto *l4proto)
+ struct nf_conntrack_tuple *tuple)
{
unsigned int size;
const __be32 *ap;
@@ -374,33 +373,20 @@ bool nf_ct_get_tuplepr(const struct sk_buff *skb, unsigned int nhoff,
u_int16_t l3num,
struct net *net, struct nf_conntrack_tuple *tuple)
{
- const struct nf_conntrack_l4proto *l4proto;
u8 protonum;
int protoff;
- int ret;
-
- rcu_read_lock();
protoff = get_l4proto(skb, nhoff, l3num, &protonum);
- if (protoff <= 0) {
- rcu_read_unlock();
+ if (protoff <= 0)
return false;
- }
- l4proto = __nf_ct_l4proto_find(protonum);
-
- ret = nf_ct_get_tuple(skb, nhoff, protoff, l3num, protonum, net, tuple,
- l4proto);
-
- rcu_read_unlock();
- return ret;
+ return nf_ct_get_tuple(skb, nhoff, protoff, l3num, protonum, net, tuple);
}
EXPORT_SYMBOL_GPL(nf_ct_get_tuplepr);
bool
nf_ct_invert_tuple(struct nf_conntrack_tuple *inverse,
- const struct nf_conntrack_tuple *orig,
- const struct nf_conntrack_l4proto *l4proto)
+ const struct nf_conntrack_tuple *orig)
{
memset(inverse, 0, sizeof(*inverse));
@@ -1354,7 +1340,6 @@ EXPORT_SYMBOL_GPL(nf_conntrack_free);
static noinline struct nf_conntrack_tuple_hash *
init_conntrack(struct net *net, struct nf_conn *tmpl,
const struct nf_conntrack_tuple *tuple,
- const struct nf_conntrack_l4proto *l4proto,
struct sk_buff *skb,
unsigned int dataoff, u32 hash)
{
@@ -1367,7 +1352,7 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
struct nf_conn_timeout *timeout_ext;
struct nf_conntrack_zone tmp;
- if (!nf_ct_invert_tuple(&repl_tuple, tuple, l4proto)) {
+ if (!nf_ct_invert_tuple(&repl_tuple, tuple)) {
pr_debug("Can't invert tuple.\n");
return NULL;
}
@@ -1449,7 +1434,6 @@ resolve_normal_ct(struct nf_conn *tmpl,
struct sk_buff *skb,
unsigned int dataoff,
u_int8_t protonum,
- const struct nf_conntrack_l4proto *l4proto,
const struct nf_hook_state *state)
{
const struct nf_conntrack_zone *zone;
@@ -1462,7 +1446,7 @@ resolve_normal_ct(struct nf_conn *tmpl,
if (!nf_ct_get_tuple(skb, skb_network_offset(skb),
dataoff, state->pf, protonum, state->net,
- &tuple, l4proto)) {
+ &tuple)) {
pr_debug("Can't get tuple\n");
return 0;
}
@@ -1472,7 +1456,7 @@ resolve_normal_ct(struct nf_conn *tmpl,
hash = hash_conntrack_raw(&tuple, state->net);
h = __nf_conntrack_find_get(state->net, zone, &tuple, hash);
if (!h) {
- h = init_conntrack(state->net, tmpl, &tuple, l4proto,
+ h = init_conntrack(state->net, tmpl, &tuple,
skb, dataoff, hash);
if (!h)
return 0;
@@ -1592,7 +1576,6 @@ static int nf_conntrack_handle_packet(struct nf_conn *ct,
unsigned int
nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
{
- const struct nf_conntrack_l4proto *l4proto;
enum ip_conntrack_info ctinfo;
struct nf_conn *ct, *tmpl;
u_int8_t protonum;
@@ -1619,8 +1602,6 @@ nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
goto out;
}
- l4proto = __nf_ct_l4proto_find(protonum);
-
if (protonum == IPPROTO_ICMP || protonum == IPPROTO_ICMPV6) {
ret = nf_conntrack_handle_icmp(tmpl, skb, dataoff,
protonum, state);
@@ -1634,7 +1615,7 @@ nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
}
repeat:
ret = resolve_normal_ct(tmpl, skb, dataoff,
- protonum, l4proto, state);
+ protonum, state);
if (ret < 0) {
/* Too stressed to deal. */
NF_CT_STAT_INC_ATOMIC(state->net, drop);
@@ -1681,19 +1662,6 @@ nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
}
EXPORT_SYMBOL_GPL(nf_conntrack_in);
-bool nf_ct_invert_tuplepr(struct nf_conntrack_tuple *inverse,
- const struct nf_conntrack_tuple *orig)
-{
- bool ret;
-
- rcu_read_lock();
- ret = nf_ct_invert_tuple(inverse, orig,
- __nf_ct_l4proto_find(orig->dst.protonum));
- rcu_read_unlock();
- return ret;
-}
-EXPORT_SYMBOL_GPL(nf_ct_invert_tuplepr);
-
/* Alter reply tuple (maybe alter helper). This is for NAT, and is
implicitly racy: see __nf_conntrack_confirm */
void nf_conntrack_alter_reply(struct nf_conn *ct,
@@ -1824,7 +1792,6 @@ static void nf_conntrack_attach(struct sk_buff *nskb, const struct sk_buff *skb)
static int nf_conntrack_update(struct net *net, struct sk_buff *skb)
{
- const struct nf_conntrack_l4proto *l4proto;
struct nf_conntrack_tuple_hash *h;
struct nf_conntrack_tuple tuple;
enum ip_conntrack_info ctinfo;
@@ -1845,10 +1812,8 @@ static int nf_conntrack_update(struct net *net, struct sk_buff *skb)
if (dataoff <= 0)
return -1;
- l4proto = nf_ct_l4proto_find_get(l4num);
-
if (!nf_ct_get_tuple(skb, skb_network_offset(skb), dataoff, l3num,
- l4num, net, &tuple, l4proto))
+ l4num, net, &tuple))
return -1;
if (ct->status & IPS_SRC_NAT) {
diff --git a/net/netfilter/nf_conntrack_pptp.c b/net/netfilter/nf_conntrack_pptp.c
index 11562f2a08bb..976f1dcb97f0 100644
--- a/net/netfilter/nf_conntrack_pptp.c
+++ b/net/netfilter/nf_conntrack_pptp.c
@@ -121,7 +121,7 @@ static void pptp_expectfn(struct nf_conn *ct,
struct nf_conntrack_expect *exp_other;
/* obviously this tuple inversion only works until you do NAT */
- nf_ct_invert_tuplepr(&inv_t, &exp->tuple);
+ nf_ct_invert_tuple(&inv_t, &exp->tuple);
pr_debug("trying to unexpect other dir: ");
nf_ct_dump_tuple(&inv_t);
diff --git a/net/netfilter/nf_conntrack_proto_icmp.c b/net/netfilter/nf_conntrack_proto_icmp.c
index d28c1d7633b2..1007efae741d 100644
--- a/net/netfilter/nf_conntrack_proto_icmp.c
+++ b/net/netfilter/nf_conntrack_proto_icmp.c
@@ -109,7 +109,6 @@ icmp_error_message(struct nf_conn *tmpl, struct sk_buff *skb,
const struct nf_hook_state *state)
{
struct nf_conntrack_tuple innertuple, origtuple;
- const struct nf_conntrack_l4proto *innerproto;
const struct nf_conntrack_tuple_hash *h;
const struct nf_conntrack_zone *zone;
enum ip_conntrack_info ctinfo;
@@ -127,12 +126,9 @@ icmp_error_message(struct nf_conn *tmpl, struct sk_buff *skb,
return -NF_ACCEPT;
}
- /* rcu_read_lock()ed by nf_hook_thresh */
- innerproto = __nf_ct_l4proto_find(origtuple.dst.protonum);
-
/* Ordinarily, we'd expect the inverted tupleproto, but it's
been preserved inside the ICMP. */
- if (!nf_ct_invert_tuple(&innertuple, &origtuple, innerproto)) {
+ if (!nf_ct_invert_tuple(&innertuple, &origtuple)) {
pr_debug("icmp_error_message: no match\n");
return -NF_ACCEPT;
}
diff --git a/net/netfilter/nf_conntrack_proto_icmpv6.c b/net/netfilter/nf_conntrack_proto_icmpv6.c
index 2910dcdea134..6c93c091a8dd 100644
--- a/net/netfilter/nf_conntrack_proto_icmpv6.c
+++ b/net/netfilter/nf_conntrack_proto_icmpv6.c
@@ -130,7 +130,6 @@ icmpv6_error_message(struct net *net, struct nf_conn *tmpl,
{
struct nf_conntrack_tuple intuple, origtuple;
const struct nf_conntrack_tuple_hash *h;
- const struct nf_conntrack_l4proto *inproto;
enum ip_conntrack_info ctinfo;
struct nf_conntrack_zone tmp;
@@ -146,12 +145,9 @@ icmpv6_error_message(struct net *net, struct nf_conn *tmpl,
return -NF_ACCEPT;
}
- /* rcu_read_lock()ed by nf_hook_thresh */
- inproto = __nf_ct_l4proto_find(origtuple.dst.protonum);
-
/* Ordinarily, we'd expect the inverted tupleproto, but it's
been preserved inside the ICMP. */
- if (!nf_ct_invert_tuple(&intuple, &origtuple, inproto)) {
+ if (!nf_ct_invert_tuple(&intuple, &origtuple)) {
pr_debug("icmpv6_error: Can't invert tuple\n");
return -NF_ACCEPT;
}
diff --git a/net/netfilter/nf_nat_core.c b/net/netfilter/nf_nat_core.c
index d159e9e7835b..44f97b3a215a 100644
--- a/net/netfilter/nf_nat_core.c
+++ b/net/netfilter/nf_nat_core.c
@@ -158,7 +158,7 @@ nf_nat_used_tuple(const struct nf_conntrack_tuple *tuple,
*/
struct nf_conntrack_tuple reply;
- nf_ct_invert_tuplepr(&reply, tuple);
+ nf_ct_invert_tuple(&reply, tuple);
return nf_conntrack_tuple_taken(&reply, ignored_conntrack);
}
EXPORT_SYMBOL(nf_nat_used_tuple);
@@ -253,7 +253,7 @@ find_appropriate_src(struct net *net,
net_eq(net, nf_ct_net(ct)) &&
nf_ct_zone_equal(ct, zone, IP_CT_DIR_ORIGINAL)) {
/* Copy source part from reply tuple. */
- nf_ct_invert_tuplepr(result,
+ nf_ct_invert_tuple(result,
&ct->tuplehash[IP_CT_DIR_REPLY].tuple);
result->dst = tuple->dst;
@@ -560,8 +560,8 @@ nf_nat_setup_info(struct nf_conn *ct,
* manipulations (future optimization: if num_manips == 0,
* orig_tp = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple)
*/
- nf_ct_invert_tuplepr(&curr_tuple,
- &ct->tuplehash[IP_CT_DIR_REPLY].tuple);
+ nf_ct_invert_tuple(&curr_tuple,
+ &ct->tuplehash[IP_CT_DIR_REPLY].tuple);
get_unique_tuple(&new_tuple, &curr_tuple, range, ct, maniptype);
@@ -569,7 +569,7 @@ nf_nat_setup_info(struct nf_conn *ct,
struct nf_conntrack_tuple reply;
/* Alter conntrack table so will recognize replies. */
- nf_ct_invert_tuplepr(&reply, &new_tuple);
+ nf_ct_invert_tuple(&reply, &new_tuple);
nf_conntrack_alter_reply(ct, &reply);
/* Non-atomic: we own this at the moment. */
@@ -640,7 +640,7 @@ static unsigned int nf_nat_manip_pkt(struct sk_buff *skb, struct nf_conn *ct,
struct nf_conntrack_tuple target;
/* We are aiming to look like inverse of other direction. */
- nf_ct_invert_tuplepr(&target, &ct->tuplehash[!dir].tuple);
+ nf_ct_invert_tuple(&target, &ct->tuplehash[!dir].tuple);
l3proto = __nf_nat_l3proto_find(target.src.l3num);
if (!l3proto->manip_pkt(skb, 0, &target, mtype))
diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c
index cd94f925495a..35884f836260 100644
--- a/net/openvswitch/conntrack.c
+++ b/net/openvswitch/conntrack.c
@@ -622,7 +622,7 @@ ovs_ct_find_existing(struct net *net, const struct nf_conntrack_zone *zone,
if (natted) {
struct nf_conntrack_tuple inverse;
- if (!nf_ct_invert_tuplepr(&inverse, &tuple)) {
+ if (!nf_ct_invert_tuple(&inverse, &tuple)) {
pr_debug("ovs_ct_find_existing: Inversion failed!\n");
return NULL;
}
--
2.11.0
^ permalink raw reply related
* [PATCH 24/33] netfilter: nft_meta: Add NFT_META_I/OIFKIND meta type
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: wenxu <wenxu@ucloud.cn>
In the ip_rcv the skb goes through the PREROUTING hook first, then kicks
in vrf device and go through the same hook again. When conntrack dnat
works with vrf, there will be some conflict with rules because the
packet goes through the hook twice with different nf status.
ip link add user1 type vrf table 1
ip link add user2 type vrf table 2
ip l set dev tun1 master user1
ip l set dev tun2 master user2
nft add table firewall
nft add chain firewall zones { type filter hook prerouting priority - 300 \; }
nft add rule firewall zones counter ct zone set iif map { "tun1" : 1, "tun2" : 2 }
nft add chain firewall rule-1000-ingress
nft add rule firewall rule-1000-ingress ct zone 1 tcp dport 22 ct state new counter accept
nft add rule firewall rule-1000-ingress counter drop
nft add chain firewall rule-1000-egress
nft add rule firewall rule-1000-egress tcp dport 22 ct state new counter drop
nft add rule firewall rule-1000-egress counter accept
nft add chain firewall rules-all { type filter hook prerouting priority - 150 \; }
nft add rule firewall rules-all ip daddr vmap { "2.2.2.11" : jump rule-1000-ingress }
nft add rule firewall rules-all ct zone vmap { 1 : jump rule-1000-egress }
nft add rule firewall dnat-all ct zone vmap { 1 : jump dnat-1000 }
nft add rule firewall dnat-1000 ip daddr 2.2.2.11 counter dnat to 10.0.0.7
For a package with ip daddr 2.2.2.11 and tcp dport 22, first time accept in the
rule-1000-ingress and dnat to 10.0.0.7. Then second time the packet goto the wrong
chain rule-1000-egress which leads the packet drop
With this patch, userspace can add the 'don't re-do entire ruleset for
vrf' policy itself via:
nft add rule firewall rules-all meta iifkind "vrf" counter accept
Signed-off-by: wenxu <wenxu@ucloud.cn>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/uapi/linux/netfilter/nf_tables.h | 4 ++++
net/netfilter/nft_meta.c | 12 ++++++++++++
2 files changed, 16 insertions(+)
diff --git a/include/uapi/linux/netfilter/nf_tables.h b/include/uapi/linux/netfilter/nf_tables.h
index 99ca95b830b6..0ba8f48bdf0b 100644
--- a/include/uapi/linux/netfilter/nf_tables.h
+++ b/include/uapi/linux/netfilter/nf_tables.h
@@ -791,6 +791,8 @@ enum nft_exthdr_attributes {
* @NFT_META_CGROUP: socket control group (skb->sk->sk_classid)
* @NFT_META_PRANDOM: a 32bit pseudo-random number
* @NFT_META_SECPATH: boolean, secpath_exists (!!skb->sp)
+ * @NFT_META_IIFKIND: packet input interface kind name (dev->rtnl_link_ops->kind)
+ * @NFT_META_OIFKIND: packet output interface kind name (dev->rtnl_link_ops->kind)
*/
enum nft_meta_keys {
NFT_META_LEN,
@@ -819,6 +821,8 @@ enum nft_meta_keys {
NFT_META_CGROUP,
NFT_META_PRANDOM,
NFT_META_SECPATH,
+ NFT_META_IIFKIND,
+ NFT_META_OIFKIND,
};
/**
diff --git a/net/netfilter/nft_meta.c b/net/netfilter/nft_meta.c
index 6df486c5ebd3..987d2d6ce624 100644
--- a/net/netfilter/nft_meta.c
+++ b/net/netfilter/nft_meta.c
@@ -244,6 +244,16 @@ void nft_meta_get_eval(const struct nft_expr *expr,
strncpy((char *)dest, p->br->dev->name, IFNAMSIZ);
return;
#endif
+ case NFT_META_IIFKIND:
+ if (in == NULL || in->rtnl_link_ops == NULL)
+ goto err;
+ strncpy((char *)dest, in->rtnl_link_ops->kind, IFNAMSIZ);
+ break;
+ case NFT_META_OIFKIND:
+ if (out == NULL || out->rtnl_link_ops == NULL)
+ goto err;
+ strncpy((char *)dest, out->rtnl_link_ops->kind, IFNAMSIZ);
+ break;
default:
WARN_ON(1);
goto err;
@@ -340,6 +350,8 @@ static int nft_meta_get_init(const struct nft_ctx *ctx,
break;
case NFT_META_IIFNAME:
case NFT_META_OIFNAME:
+ case NFT_META_IIFKIND:
+ case NFT_META_OIFKIND:
len = IFNAMSIZ;
break;
case NFT_META_PRANDOM:
--
2.11.0
^ permalink raw reply related
* [PATCH 28/33] netfilter: conntrack: fix bogus port values for other l4 protocols
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
We must only extract l4 proto information if we can track the layer 4
protocol.
Before removal of pkt_to_tuple callback, the code to extract port
information was only reached for TCP/UDP/LITE/DCCP/SCTP.
The other protocols were handled by the indirect call, and the
'generic' tracker took care of other protocols that have no notion
of 'ports'.
After removal of the callback we must be more strict here and only
init port numbers for those protocols that have ports.
Fixes: df5e1629087a ("netfilter: conntrack: remove pkt_to_tuple callback")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
net/netfilter/nf_conntrack_core.c | 46 +++++++++++++++++++++++++++++----------
1 file changed, 35 insertions(+), 11 deletions(-)
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index a3e5232c2088..815956ac5a76 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -222,6 +222,24 @@ static u32 hash_conntrack(const struct net *net,
return scale_hash(hash_conntrack_raw(tuple, net));
}
+static bool nf_ct_get_tuple_ports(const struct sk_buff *skb,
+ unsigned int dataoff,
+ struct nf_conntrack_tuple *tuple)
+{ struct {
+ __be16 sport;
+ __be16 dport;
+ } _inet_hdr, *inet_hdr;
+
+ /* Actually only need first 4 bytes to get ports. */
+ inet_hdr = skb_header_pointer(skb, dataoff, sizeof(_inet_hdr), &_inet_hdr);
+ if (!inet_hdr)
+ return false;
+
+ tuple->src.u.udp.port = inet_hdr->sport;
+ tuple->dst.u.udp.port = inet_hdr->dport;
+ return true;
+}
+
static bool
nf_ct_get_tuple(const struct sk_buff *skb,
unsigned int nhoff,
@@ -234,10 +252,6 @@ nf_ct_get_tuple(const struct sk_buff *skb,
unsigned int size;
const __be32 *ap;
__be32 _addrs[8];
- struct {
- __be16 sport;
- __be16 dport;
- } _inet_hdr, *inet_hdr;
memset(tuple, 0, sizeof(*tuple));
@@ -284,15 +298,25 @@ nf_ct_get_tuple(const struct sk_buff *skb,
case IPPROTO_GRE:
return gre_pkt_to_tuple(skb, dataoff, net, tuple);
#endif
+ case IPPROTO_TCP:
+ case IPPROTO_UDP: /* fallthrough */
+ return nf_ct_get_tuple_ports(skb, dataoff, tuple);
+#ifdef CONFIG_NF_CT_PROTO_UDPLITE
+ case IPPROTO_UDPLITE:
+ return nf_ct_get_tuple_ports(skb, dataoff, tuple);
+#endif
+#ifdef CONFIG_NF_CT_PROTO_SCTP
+ case IPPROTO_SCTP:
+ return nf_ct_get_tuple_ports(skb, dataoff, tuple);
+#endif
+#ifdef CONFIG_NF_CT_PROTO_DCCP
+ case IPPROTO_DCCP:
+ return nf_ct_get_tuple_ports(skb, dataoff, tuple);
+#endif
+ default:
+ break;
}
- /* Actually only need first 4 bytes to get ports. */
- inet_hdr = skb_header_pointer(skb, dataoff, sizeof(_inet_hdr), &_inet_hdr);
- if (!inet_hdr)
- return false;
-
- tuple->src.u.udp.port = inet_hdr->sport;
- tuple->dst.u.udp.port = inet_hdr->dport;
return true;
}
--
2.11.0
^ permalink raw reply related
* [PATCH 25/33] netfilter: nat: un-export nf_nat_used_tuple
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
Not used since 203f2e78200c27e ("netfilter: nat: remove l4proto->unique_tuple")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_nat.h | 4 ----
net/netfilter/nf_nat_core.c | 3 +--
2 files changed, 1 insertion(+), 6 deletions(-)
diff --git a/include/net/netfilter/nf_nat.h b/include/net/netfilter/nf_nat.h
index a17eb2f8d40e..8aff77cafb8b 100644
--- a/include/net/netfilter/nf_nat.h
+++ b/include/net/netfilter/nf_nat.h
@@ -47,10 +47,6 @@ extern unsigned int nf_nat_alloc_null_binding(struct nf_conn *ct,
struct nf_conn_nat *nf_ct_nat_ext_add(struct nf_conn *ct);
-/* Is this tuple already taken? (not by us)*/
-int nf_nat_used_tuple(const struct nf_conntrack_tuple *tuple,
- const struct nf_conn *ignored_conntrack);
-
static inline struct nf_conn_nat *nfct_nat(const struct nf_conn *ct)
{
#if defined(CONFIG_NF_NAT) || defined(CONFIG_NF_NAT_MODULE)
diff --git a/net/netfilter/nf_nat_core.c b/net/netfilter/nf_nat_core.c
index 44f97b3a215a..35e61038ae96 100644
--- a/net/netfilter/nf_nat_core.c
+++ b/net/netfilter/nf_nat_core.c
@@ -146,7 +146,7 @@ hash_by_src(const struct net *n, const struct nf_conntrack_tuple *tuple)
}
/* Is this tuple already taken? (not by us) */
-int
+static int
nf_nat_used_tuple(const struct nf_conntrack_tuple *tuple,
const struct nf_conn *ignored_conntrack)
{
@@ -161,7 +161,6 @@ nf_nat_used_tuple(const struct nf_conntrack_tuple *tuple,
nf_ct_invert_tuple(&reply, tuple);
return nf_conntrack_tuple_taken(&reply, ignored_conntrack);
}
-EXPORT_SYMBOL(nf_nat_used_tuple);
static bool nf_nat_inet_in_range(const struct nf_conntrack_tuple *t,
const struct nf_nat_range2 *range)
--
2.11.0
^ permalink raw reply related
* [PATCH 22/33] netfilter: conntrack: remove nf_ct_l4proto_find_get
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
Its now same as __nf_ct_l4proto_find(), so rename that to
nf_ct_l4proto_find and use it everywhere.
It never returns NULL and doesn't need locks or reference counts.
Before this series:
302824 net/netfilter/nf_conntrack.ko
21504 net/netfilter/nf_conntrack_proto_gre.ko
text data bss dec hex filename
6281 1732 4 8017 1f51 nf_conntrack_proto_gre.ko
108356 20613 236 129205 1f8b5 nf_conntrack.ko
After:
294864 net/netfilter/nf_conntrack.ko
text data bss dec hex filename
106979 19557 240 126776 1ef38 nf_conntrack.ko
so, even with builtin gre, total size got reduced.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/net/netfilter/nf_conntrack_core.h | 2 +-
include/net/netfilter/nf_conntrack_l4proto.h | 8 +-
net/netfilter/nf_conntrack_core.c | 11 +-
net/netfilter/nf_conntrack_expect.c | 2 +-
net/netfilter/nf_conntrack_netlink.c | 14 +--
net/netfilter/nf_conntrack_proto.c | 175 ++++-----------------------
net/netfilter/nf_conntrack_standalone.c | 3 +-
net/netfilter/nf_flow_table_core.c | 2 +-
net/netfilter/nfnetlink_cttimeout.c | 6 +-
net/netfilter/nft_ct.c | 2 +-
net/netfilter/xt_CT.c | 2 +-
11 files changed, 45 insertions(+), 182 deletions(-)
diff --git a/include/net/netfilter/nf_conntrack_core.h b/include/net/netfilter/nf_conntrack_core.h
index 235c182022b2..ae41e92251dd 100644
--- a/include/net/netfilter/nf_conntrack_core.h
+++ b/include/net/netfilter/nf_conntrack_core.h
@@ -26,7 +26,7 @@ int nf_conntrack_init_net(struct net *net);
void nf_conntrack_cleanup_net(struct net *net);
void nf_conntrack_cleanup_net_list(struct list_head *net_exit_list);
-int nf_conntrack_proto_pernet_init(struct net *net);
+void nf_conntrack_proto_pernet_init(struct net *net);
void nf_conntrack_proto_pernet_fini(struct net *net);
int nf_conntrack_proto_init(void);
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index 5d1419ac6a38..778087591983 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -140,13 +140,7 @@ extern const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic;
#define MAX_NF_CT_PROTO IPPROTO_UDPLITE
-const struct nf_conntrack_l4proto *__nf_ct_l4proto_find(u8 l4proto);
-
-const struct nf_conntrack_l4proto *nf_ct_l4proto_find_get(u8 l4proto);
-
-/* Protocol global registration. */
-int nf_ct_l4proto_register_one(const struct nf_conntrack_l4proto *proto);
-void nf_ct_l4proto_unregister_one(const struct nf_conntrack_l4proto *proto);
+const struct nf_conntrack_l4proto *nf_ct_l4proto_find(u8 l4proto);
/* Generic netlink helpers */
int nf_ct_port_tuple_to_nlattr(struct sk_buff *skb,
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 52e6c5c6f202..171659aa69a1 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -845,7 +845,7 @@ static int nf_ct_resolve_clash(struct net *net, struct sk_buff *skb,
enum ip_conntrack_info oldinfo;
struct nf_conn *loser_ct = nf_ct_get(skb, &oldinfo);
- l4proto = __nf_ct_l4proto_find(nf_ct_protonum(ct));
+ l4proto = nf_ct_l4proto_find(nf_ct_protonum(ct));
if (l4proto->allow_clash &&
!nf_ct_is_dying(ct) &&
atomic_inc_not_zero(&ct->ct_general.use)) {
@@ -1117,7 +1117,7 @@ static bool gc_worker_can_early_drop(const struct nf_conn *ct)
if (!test_bit(IPS_ASSURED_BIT, &ct->status))
return true;
- l4proto = __nf_ct_l4proto_find(nf_ct_protonum(ct));
+ l4proto = nf_ct_l4proto_find(nf_ct_protonum(ct));
if (l4proto->can_early_drop && l4proto->can_early_drop(ct))
return true;
@@ -2452,15 +2452,10 @@ int nf_conntrack_init_net(struct net *net)
nf_conntrack_tstamp_pernet_init(net);
nf_conntrack_ecache_pernet_init(net);
nf_conntrack_helper_pernet_init(net);
+ nf_conntrack_proto_pernet_init(net);
- ret = nf_conntrack_proto_pernet_init(net);
- if (ret < 0)
- goto err_proto;
return 0;
-err_proto:
- nf_conntrack_ecache_pernet_fini(net);
- nf_conntrack_expect_pernet_fini(net);
err_expect:
free_percpu(net->ct.stat);
err_pcpu_lists:
diff --git a/net/netfilter/nf_conntrack_expect.c b/net/netfilter/nf_conntrack_expect.c
index 3034038bfdf0..334d6e5b7762 100644
--- a/net/netfilter/nf_conntrack_expect.c
+++ b/net/netfilter/nf_conntrack_expect.c
@@ -610,7 +610,7 @@ static int exp_seq_show(struct seq_file *s, void *v)
expect->tuple.src.l3num,
expect->tuple.dst.protonum);
print_tuple(s, &expect->tuple,
- __nf_ct_l4proto_find(expect->tuple.dst.protonum));
+ nf_ct_l4proto_find(expect->tuple.dst.protonum));
if (expect->flags & NF_CT_EXPECT_PERMANENT) {
seq_puts(s, "PERMANENT");
diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c
index 1213beb5a714..8071bb04a849 100644
--- a/net/netfilter/nf_conntrack_netlink.c
+++ b/net/netfilter/nf_conntrack_netlink.c
@@ -134,7 +134,7 @@ static int ctnetlink_dump_tuples(struct sk_buff *skb,
ret = ctnetlink_dump_tuples_ip(skb, tuple);
if (ret >= 0) {
- l4proto = __nf_ct_l4proto_find(tuple->dst.protonum);
+ l4proto = nf_ct_l4proto_find(tuple->dst.protonum);
ret = ctnetlink_dump_tuples_proto(skb, tuple, l4proto);
}
rcu_read_unlock();
@@ -182,7 +182,7 @@ static int ctnetlink_dump_protoinfo(struct sk_buff *skb, struct nf_conn *ct)
struct nlattr *nest_proto;
int ret;
- l4proto = __nf_ct_l4proto_find(nf_ct_protonum(ct));
+ l4proto = nf_ct_l4proto_find(nf_ct_protonum(ct));
if (!l4proto->to_nlattr)
return 0;
@@ -590,7 +590,7 @@ static size_t ctnetlink_proto_size(const struct nf_conn *ct)
len = nla_policy_len(cta_ip_nla_policy, CTA_IP_MAX + 1);
len *= 3u; /* ORIG, REPLY, MASTER */
- l4proto = __nf_ct_l4proto_find(nf_ct_protonum(ct));
+ l4proto = nf_ct_l4proto_find(nf_ct_protonum(ct));
len += l4proto->nlattr_size;
if (l4proto->nlattr_tuple_size) {
len4 = l4proto->nlattr_tuple_size();
@@ -1059,7 +1059,7 @@ static int ctnetlink_parse_tuple_proto(struct nlattr *attr,
tuple->dst.protonum = nla_get_u8(tb[CTA_PROTO_NUM]);
rcu_read_lock();
- l4proto = __nf_ct_l4proto_find(tuple->dst.protonum);
+ l4proto = nf_ct_l4proto_find(tuple->dst.protonum);
if (likely(l4proto->nlattr_to_tuple)) {
ret = nla_validate_nested(attr, CTA_PROTO_MAX,
@@ -1722,11 +1722,9 @@ static int ctnetlink_change_protoinfo(struct nf_conn *ct,
if (err < 0)
return err;
- rcu_read_lock();
- l4proto = __nf_ct_l4proto_find(nf_ct_protonum(ct));
+ l4proto = nf_ct_l4proto_find(nf_ct_protonum(ct));
if (l4proto->from_nlattr)
err = l4proto->from_nlattr(tb, ct);
- rcu_read_unlock();
return err;
}
@@ -2676,7 +2674,7 @@ static int ctnetlink_exp_dump_mask(struct sk_buff *skb,
rcu_read_lock();
ret = ctnetlink_dump_tuples_ip(skb, &m);
if (ret >= 0) {
- l4proto = __nf_ct_l4proto_find(tuple->dst.protonum);
+ l4proto = nf_ct_l4proto_find(tuple->dst.protonum);
ret = ctnetlink_dump_tuples_proto(skb, &m, l4proto);
}
rcu_read_unlock();
diff --git a/net/netfilter/nf_conntrack_proto.c b/net/netfilter/nf_conntrack_proto.c
index e6bc02c13f0f..aa8d3fe0b37f 100644
--- a/net/netfilter/nf_conntrack_proto.c
+++ b/net/netfilter/nf_conntrack_proto.c
@@ -43,8 +43,6 @@
extern unsigned int nf_conntrack_net_id;
-static struct nf_conntrack_l4proto __rcu *nf_ct_protos[MAX_NF_CT_PROTO + 1] __read_mostly;
-
static DEFINE_MUTEX(nf_ct_proto_mutex);
#ifdef CONFIG_SYSCTL
@@ -95,121 +93,32 @@ void nf_ct_l4proto_log_invalid(const struct sk_buff *skb,
EXPORT_SYMBOL_GPL(nf_ct_l4proto_log_invalid);
#endif
-const struct nf_conntrack_l4proto *__nf_ct_l4proto_find(u8 l4proto)
-{
- if (unlikely(l4proto >= ARRAY_SIZE(nf_ct_protos)))
- return &nf_conntrack_l4proto_generic;
-
- return rcu_dereference(nf_ct_protos[l4proto]);
-}
-EXPORT_SYMBOL_GPL(__nf_ct_l4proto_find);
-
-const struct nf_conntrack_l4proto *nf_ct_l4proto_find_get(u8 l4num)
-{
- const struct nf_conntrack_l4proto *p;
-
- rcu_read_lock();
- p = __nf_ct_l4proto_find(l4num);
- rcu_read_unlock();
-
- return p;
-}
-EXPORT_SYMBOL_GPL(nf_ct_l4proto_find_get);
-
-static int kill_l4proto(struct nf_conn *i, void *data)
+const struct nf_conntrack_l4proto *nf_ct_l4proto_find(u8 l4proto)
{
- const struct nf_conntrack_l4proto *l4proto;
- l4proto = data;
- return nf_ct_protonum(i) == l4proto->l4proto;
-}
-
-/* FIXME: Allow NULL functions and sub in pointers to generic for
- them. --RR */
-int nf_ct_l4proto_register_one(const struct nf_conntrack_l4proto *l4proto)
-{
- int ret = 0;
-
- if ((l4proto->to_nlattr && l4proto->nlattr_size == 0) ||
- (l4proto->tuple_to_nlattr && !l4proto->nlattr_tuple_size))
- return -EINVAL;
-
- mutex_lock(&nf_ct_proto_mutex);
- if (rcu_dereference_protected(
- nf_ct_protos[l4proto->l4proto],
- lockdep_is_held(&nf_ct_proto_mutex)
- ) != &nf_conntrack_l4proto_generic) {
- ret = -EBUSY;
- goto out_unlock;
+ switch (l4proto) {
+ case IPPROTO_UDP: return &nf_conntrack_l4proto_udp;
+ case IPPROTO_TCP: return &nf_conntrack_l4proto_tcp;
+ case IPPROTO_ICMP: return &nf_conntrack_l4proto_icmp;
+#ifdef CONFIG_NF_CT_PROTO_DCCP
+ case IPPROTO_DCCP: return &nf_conntrack_l4proto_dccp;
+#endif
+#ifdef CONFIG_NF_CT_PROTO_SCTP
+ case IPPROTO_SCTP: return &nf_conntrack_l4proto_sctp;
+#endif
+#ifdef CONFIG_NF_CT_PROTO_UDPLITE
+ case IPPROTO_UDPLITE: return &nf_conntrack_l4proto_udplite;
+#endif
+#ifdef CONFIG_NF_CT_PROTO_GRE
+ case IPPROTO_GRE: return &nf_conntrack_l4proto_gre;
+#endif
+#if IS_ENABLED(CONFIG_IPV6)
+ case IPPROTO_ICMPV6: return &nf_conntrack_l4proto_icmpv6;
+#endif /* CONFIG_IPV6 */
}
- rcu_assign_pointer(nf_ct_protos[l4proto->l4proto], l4proto);
-out_unlock:
- mutex_unlock(&nf_ct_proto_mutex);
- return ret;
-}
-EXPORT_SYMBOL_GPL(nf_ct_l4proto_register_one);
-
-static void __nf_ct_l4proto_unregister_one(const struct nf_conntrack_l4proto *l4proto)
-
-{
- BUG_ON(l4proto->l4proto >= ARRAY_SIZE(nf_ct_protos));
-
- BUG_ON(rcu_dereference_protected(
- nf_ct_protos[l4proto->l4proto],
- lockdep_is_held(&nf_ct_proto_mutex)
- ) != l4proto);
- rcu_assign_pointer(nf_ct_protos[l4proto->l4proto],
- &nf_conntrack_l4proto_generic);
-}
-
-void nf_ct_l4proto_unregister_one(const struct nf_conntrack_l4proto *l4proto)
-{
- mutex_lock(&nf_ct_proto_mutex);
- __nf_ct_l4proto_unregister_one(l4proto);
- mutex_unlock(&nf_ct_proto_mutex);
-
- synchronize_net();
- /* Remove all contrack entries for this protocol */
- nf_ct_iterate_destroy(kill_l4proto, (void *)l4proto);
-}
-EXPORT_SYMBOL_GPL(nf_ct_l4proto_unregister_one);
-
-static void
-nf_ct_l4proto_unregister(const struct nf_conntrack_l4proto * const l4proto[],
- unsigned int num_proto)
-{
- int i;
-
- mutex_lock(&nf_ct_proto_mutex);
- for (i = 0; i < num_proto; i++)
- __nf_ct_l4proto_unregister_one(l4proto[i]);
- mutex_unlock(&nf_ct_proto_mutex);
-
- synchronize_net();
-
- for (i = 0; i < num_proto; i++)
- nf_ct_iterate_destroy(kill_l4proto, (void *)l4proto[i]);
-}
-
-static int
-nf_ct_l4proto_register(const struct nf_conntrack_l4proto * const l4proto[],
- unsigned int num_proto)
-{
- int ret = -EINVAL;
- unsigned int i;
-
- for (i = 0; i < num_proto; i++) {
- ret = nf_ct_l4proto_register_one(l4proto[i]);
- if (ret < 0)
- break;
- }
- if (i != num_proto) {
- pr_err("nf_conntrack: can't register l4 %d proto.\n",
- l4proto[i]->l4proto);
- nf_ct_l4proto_unregister(l4proto, i);
- }
- return ret;
-}
+ return &nf_conntrack_l4proto_generic;
+};
+EXPORT_SYMBOL_GPL(nf_ct_l4proto_find);
static unsigned int nf_confirm(struct sk_buff *skb,
unsigned int protoff,
@@ -651,30 +560,9 @@ void nf_ct_netns_put(struct net *net, uint8_t nfproto)
}
EXPORT_SYMBOL_GPL(nf_ct_netns_put);
-static const struct nf_conntrack_l4proto * const builtin_l4proto[] = {
- &nf_conntrack_l4proto_tcp,
- &nf_conntrack_l4proto_udp,
- &nf_conntrack_l4proto_icmp,
-#ifdef CONFIG_NF_CT_PROTO_DCCP
- &nf_conntrack_l4proto_dccp,
-#endif
-#ifdef CONFIG_NF_CT_PROTO_SCTP
- &nf_conntrack_l4proto_sctp,
-#endif
-#ifdef CONFIG_NF_CT_PROTO_UDPLITE
- &nf_conntrack_l4proto_udplite,
-#endif
-#ifdef CONFIG_NF_CT_PROTO_GRE
- &nf_conntrack_l4proto_gre,
-#endif
-#if IS_ENABLED(CONFIG_IPV6)
- &nf_conntrack_l4proto_icmpv6,
-#endif /* CONFIG_IPV6 */
-};
-
int nf_conntrack_proto_init(void)
{
- int ret = 0, i;
+ int ret;
ret = nf_register_sockopt(&so_getorigdst);
if (ret < 0)
@@ -686,18 +574,8 @@ int nf_conntrack_proto_init(void)
goto cleanup_sockopt;
#endif
- for (i = 0; i < ARRAY_SIZE(nf_ct_protos); i++)
- RCU_INIT_POINTER(nf_ct_protos[i],
- &nf_conntrack_l4proto_generic);
-
- ret = nf_ct_l4proto_register(builtin_l4proto,
- ARRAY_SIZE(builtin_l4proto));
- if (ret < 0)
- goto cleanup_sockopt2;
-
return ret;
-cleanup_sockopt2:
- nf_unregister_sockopt(&so_getorigdst);
+
#if IS_ENABLED(CONFIG_IPV6)
cleanup_sockopt:
nf_unregister_sockopt(&so_getorigdst6);
@@ -713,7 +591,7 @@ void nf_conntrack_proto_fini(void)
#endif
}
-int nf_conntrack_proto_pernet_init(struct net *net)
+void nf_conntrack_proto_pernet_init(struct net *net)
{
nf_conntrack_generic_init_net(net);
nf_conntrack_udp_init_net(net);
@@ -729,7 +607,6 @@ int nf_conntrack_proto_pernet_init(struct net *net)
#ifdef CONFIG_NF_CT_PROTO_GRE
nf_conntrack_gre_init_net(net);
#endif
- return 0;
}
void nf_conntrack_proto_pernet_fini(struct net *net)
diff --git a/net/netfilter/nf_conntrack_standalone.c b/net/netfilter/nf_conntrack_standalone.c
index d848de713dc0..ddfca5f1784c 100644
--- a/net/netfilter/nf_conntrack_standalone.c
+++ b/net/netfilter/nf_conntrack_standalone.c
@@ -310,8 +310,7 @@ static int ct_seq_show(struct seq_file *s, void *v)
if (!net_eq(nf_ct_net(ct), net))
goto release;
- l4proto = __nf_ct_l4proto_find(nf_ct_protonum(ct));
- WARN_ON(!l4proto);
+ l4proto = nf_ct_l4proto_find(nf_ct_protonum(ct));
ret = -ENOSPC;
seq_printf(s, "%-8s %u %-8s %u ",
diff --git a/net/netfilter/nf_flow_table_core.c b/net/netfilter/nf_flow_table_core.c
index fa0844e2a68d..8099f0f778ab 100644
--- a/net/netfilter/nf_flow_table_core.c
+++ b/net/netfilter/nf_flow_table_core.c
@@ -120,7 +120,7 @@ static void flow_offload_fixup_ct_state(struct nf_conn *ct)
if (l4num == IPPROTO_TCP)
flow_offload_fixup_tcp(&ct->proto.tcp);
- l4proto = __nf_ct_l4proto_find(l4num);
+ l4proto = nf_ct_l4proto_find(l4num);
if (!l4proto)
return;
diff --git a/net/netfilter/nfnetlink_cttimeout.c b/net/netfilter/nfnetlink_cttimeout.c
index 37b4f84ac153..c69b11ca5aad 100644
--- a/net/netfilter/nfnetlink_cttimeout.c
+++ b/net/netfilter/nfnetlink_cttimeout.c
@@ -122,7 +122,7 @@ static int cttimeout_new_timeout(struct net *net, struct sock *ctnl,
return -EBUSY;
}
- l4proto = nf_ct_l4proto_find_get(l4num);
+ l4proto = nf_ct_l4proto_find(l4num);
/* This protocol is not supportted, skip. */
if (l4proto->l4proto != l4num) {
@@ -357,7 +357,7 @@ static int cttimeout_default_set(struct net *net, struct sock *ctnl,
return -EINVAL;
l4num = nla_get_u8(cda[CTA_TIMEOUT_L4PROTO]);
- l4proto = nf_ct_l4proto_find_get(l4num);
+ l4proto = nf_ct_l4proto_find(l4num);
/* This protocol is not supported, skip. */
if (l4proto->l4proto != l4num) {
@@ -438,7 +438,7 @@ static int cttimeout_default_get(struct net *net, struct sock *ctnl,
l3num = ntohs(nla_get_be16(cda[CTA_TIMEOUT_L3PROTO]));
l4num = nla_get_u8(cda[CTA_TIMEOUT_L4PROTO]);
- l4proto = nf_ct_l4proto_find_get(l4num);
+ l4proto = nf_ct_l4proto_find(l4num);
err = -EOPNOTSUPP;
if (l4proto->l4proto != l4num)
diff --git a/net/netfilter/nft_ct.c b/net/netfilter/nft_ct.c
index 3249cc059048..7b717fad6cdc 100644
--- a/net/netfilter/nft_ct.c
+++ b/net/netfilter/nft_ct.c
@@ -870,7 +870,7 @@ static int nft_ct_timeout_obj_init(const struct nft_ctx *ctx,
l4num = nla_get_u8(tb[NFTA_CT_TIMEOUT_L4PROTO]);
priv->l4proto = l4num;
- l4proto = nf_ct_l4proto_find_get(l4num);
+ l4proto = nf_ct_l4proto_find(l4num);
if (l4proto->l4proto != l4num) {
ret = -EOPNOTSUPP;
diff --git a/net/netfilter/xt_CT.c b/net/netfilter/xt_CT.c
index 2c7a4b80206f..0fa863f57575 100644
--- a/net/netfilter/xt_CT.c
+++ b/net/netfilter/xt_CT.c
@@ -159,7 +159,7 @@ xt_ct_set_timeout(struct nf_conn *ct, const struct xt_tgchk_param *par,
/* Make sure the timeout policy matches any existing protocol tracker,
* otherwise default to generic.
*/
- l4proto = __nf_ct_l4proto_find(proto);
+ l4proto = nf_ct_l4proto_find(proto);
if (timeout->l4proto->l4proto != l4proto->l4proto) {
ret = -EINVAL;
pr_info_ratelimited("Timeout policy `%s' can only be used by L%d protocol number %d\n",
--
2.11.0
^ permalink raw reply related
* [PATCH 32/33] netfilter: conntrack: fix error path in nf_conntrack_pernet_init()
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Cong Wang <xiyou.wangcong@gmail.com>
When nf_ct_netns_get() fails, it should clean up itself,
its caller doesn't need to call nf_conntrack_fini_net().
nf_conntrack_init_net() is called after registering sysctl
and proc, so its cleanup function should be called before
unregistering sysctl and proc.
Fixes: ba3fbe663635 ("netfilter: nf_conntrack: provide modparam to always register conntrack hooks")
Fixes: b884fa461776 ("netfilter: conntrack: unify sysctl handling")
Reported-and-tested-by: syzbot+fcee88b2d87f0539dfe9@syzkaller.appspotmail.com
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
net/netfilter/nf_conntrack_standalone.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/netfilter/nf_conntrack_standalone.c b/net/netfilter/nf_conntrack_standalone.c
index 8928a4d0933e..c2ae14c720b4 100644
--- a/net/netfilter/nf_conntrack_standalone.c
+++ b/net/netfilter/nf_conntrack_standalone.c
@@ -1115,11 +1115,11 @@ static int nf_conntrack_pernet_init(struct net *net)
return 0;
out_hooks:
- nf_conntrack_fini_net(net);
+ nf_conntrack_cleanup_net(net);
out_init_net:
nf_conntrack_standalone_fini_proc(net);
out_proc:
- nf_conntrack_cleanup_net(net);
+ nf_conntrack_standalone_fini_sysctl(net);
return ret;
}
--
2.11.0
^ permalink raw reply related
* [PATCH 33/33] netfilter: ipv4: remove useless export_symbol
From: Pablo Neira Ayuso @ 2019-01-28 23:57 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
In-Reply-To: <20190128235750.18412-1-pablo@netfilter.org>
From: Florian Westphal <fw@strlen.de>
Only one caller; place it where needed and get rid of the EXPORT_SYMBOL.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
include/linux/netfilter_ipv4.h | 6 ------
net/ipv4/netfilter.c | 18 ------------------
net/netfilter/utils.c | 19 +++++++++++++++++++
3 files changed, 19 insertions(+), 24 deletions(-)
diff --git a/include/linux/netfilter_ipv4.h b/include/linux/netfilter_ipv4.h
index 95ab5cc64422..082e2c41b7ff 100644
--- a/include/linux/netfilter_ipv4.h
+++ b/include/linux/netfilter_ipv4.h
@@ -25,7 +25,6 @@ __sum16 nf_ip_checksum(struct sk_buff *skb, unsigned int hook,
unsigned int dataoff, u_int8_t protocol);
int nf_ip_route(struct net *net, struct dst_entry **dst, struct flowi *fl,
bool strict);
-int nf_ip_reroute(struct sk_buff *skb, const struct nf_queue_entry *entry);
#else
static inline __sum16 nf_ip_checksum(struct sk_buff *skb, unsigned int hook,
unsigned int dataoff, u_int8_t protocol)
@@ -37,11 +36,6 @@ static inline int nf_ip_route(struct net *net, struct dst_entry **dst,
{
return -EOPNOTSUPP;
}
-static inline int nf_ip_reroute(struct sk_buff *skb,
- const struct nf_queue_entry *entry)
-{
- return -EOPNOTSUPP;
-}
#endif /* CONFIG_INET */
#endif /*__LINUX_IP_NETFILTER_H*/
diff --git a/net/ipv4/netfilter.c b/net/ipv4/netfilter.c
index 8d2e5dc9a827..a058213b77a7 100644
--- a/net/ipv4/netfilter.c
+++ b/net/ipv4/netfilter.c
@@ -80,24 +80,6 @@ int ip_route_me_harder(struct net *net, struct sk_buff *skb, unsigned int addr_t
}
EXPORT_SYMBOL(ip_route_me_harder);
-int nf_ip_reroute(struct sk_buff *skb, const struct nf_queue_entry *entry)
-{
- const struct ip_rt_info *rt_info = nf_queue_entry_reroute(entry);
-
- if (entry->state.hook == NF_INET_LOCAL_OUT) {
- const struct iphdr *iph = ip_hdr(skb);
-
- if (!(iph->tos == rt_info->tos &&
- skb->mark == rt_info->mark &&
- iph->daddr == rt_info->daddr &&
- iph->saddr == rt_info->saddr))
- return ip_route_me_harder(entry->state.net, skb,
- RTN_UNSPEC);
- }
- return 0;
-}
-EXPORT_SYMBOL_GPL(nf_ip_reroute);
-
int nf_ip_route(struct net *net, struct dst_entry **dst, struct flowi *fl,
bool strict __always_unused)
{
diff --git a/net/netfilter/utils.c b/net/netfilter/utils.c
index e8da9a9bba73..55af9f247993 100644
--- a/net/netfilter/utils.c
+++ b/net/netfilter/utils.c
@@ -180,6 +180,25 @@ int nf_route(struct net *net, struct dst_entry **dst, struct flowi *fl,
}
EXPORT_SYMBOL_GPL(nf_route);
+static int nf_ip_reroute(struct sk_buff *skb, const struct nf_queue_entry *entry)
+{
+#ifdef CONFIG_INET
+ const struct ip_rt_info *rt_info = nf_queue_entry_reroute(entry);
+
+ if (entry->state.hook == NF_INET_LOCAL_OUT) {
+ const struct iphdr *iph = ip_hdr(skb);
+
+ if (!(iph->tos == rt_info->tos &&
+ skb->mark == rt_info->mark &&
+ iph->daddr == rt_info->daddr &&
+ iph->saddr == rt_info->saddr))
+ return ip_route_me_harder(entry->state.net, skb,
+ RTN_UNSPEC);
+ }
+#endif
+ return 0;
+}
+
int nf_reroute(struct sk_buff *skb, struct nf_queue_entry *entry)
{
const struct nf_ipv6_ops *v6ops;
--
2.11.0
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox