summaryrefslogtreecommitdiff
path: root/net
diff options
context:
space:
mode:
authorDavid S. Miller <davem@davemloft.net>2020-02-29 15:53:35 -0800
committerDavid S. Miller <davem@davemloft.net>2020-02-29 15:53:35 -0800
commit9f0ca0c1a50a7de5c71970aa452941199ed210d9 (patch)
treeb7e19ab2e5bbc951141fb90489923b50b8f263f3 /net
parent9a834f9b04bc5ed2db27eb45b0cc6415ce7c314c (diff)
parent812285fa5ab129e3a55682314bf235f701564310 (diff)
Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next
Alexei Starovoitov says: ==================== pull-request: bpf-next 2020-02-28 The following pull-request contains BPF updates for your *net-next* tree. We've added 41 non-merge commits during the last 7 day(s) which contain a total of 49 files changed, 1383 insertions(+), 499 deletions(-). The main changes are: 1) BPF and Real-Time nicely co-exist. 2) bpftool feature improvements. 3) retrieve bpf_sk_storage via INET_DIAG. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net')
-rw-r--r--net/bpf/test_run.c8
-rw-r--r--net/core/bpf_sk_storage.c283
-rw-r--r--net/core/flow_dissector.c4
-rw-r--r--net/core/skmsg.c8
-rw-r--r--net/dccp/diag.c9
-rw-r--r--net/ipv4/inet_diag.c307
-rw-r--r--net/ipv4/raw_diag.c24
-rw-r--r--net/ipv4/tcp_diag.c8
-rw-r--r--net/ipv4/udp_diag.c41
-rw-r--r--net/kcm/kcmsock.c4
-rw-r--r--net/sctp/diag.c7
11 files changed, 516 insertions, 187 deletions
diff --git a/net/bpf/test_run.c b/net/bpf/test_run.c
index d555c0d8657d..562443f94133 100644
--- a/net/bpf/test_run.c
+++ b/net/bpf/test_run.c
@@ -37,7 +37,7 @@ static int bpf_test_run(struct bpf_prog *prog, void *ctx, u32 repeat,
repeat = 1;
rcu_read_lock();
- preempt_disable();
+ migrate_disable();
time_start = ktime_get_ns();
for (i = 0; i < repeat; i++) {
bpf_cgroup_storage_set(storage);
@@ -54,18 +54,18 @@ static int bpf_test_run(struct bpf_prog *prog, void *ctx, u32 repeat,
if (need_resched()) {
time_spent += ktime_get_ns() - time_start;
- preempt_enable();
+ migrate_enable();
rcu_read_unlock();
cond_resched();
rcu_read_lock();
- preempt_disable();
+ migrate_disable();
time_start = ktime_get_ns();
}
}
time_spent += ktime_get_ns() - time_start;
- preempt_enable();
+ migrate_enable();
rcu_read_unlock();
do_div(time_spent, repeat);
diff --git a/net/core/bpf_sk_storage.c b/net/core/bpf_sk_storage.c
index 427cfbc0d50d..756b63b6f7b3 100644
--- a/net/core/bpf_sk_storage.c
+++ b/net/core/bpf_sk_storage.c
@@ -8,6 +8,7 @@
#include <linux/bpf.h>
#include <net/bpf_sk_storage.h>
#include <net/sock.h>
+#include <uapi/linux/sock_diag.h>
#include <uapi/linux/btf.h>
static atomic_t cache_idx;
@@ -606,6 +607,14 @@ static void bpf_sk_storage_map_free(struct bpf_map *map)
kfree(map);
}
+/* U16_MAX is much more than enough for sk local storage
+ * considering a tcp_sock is ~2k.
+ */
+#define MAX_VALUE_SIZE \
+ min_t(u32, \
+ (KMALLOC_MAX_SIZE - MAX_BPF_STACK - sizeof(struct bpf_sk_storage_elem)), \
+ (U16_MAX - sizeof(struct bpf_sk_storage_elem)))
+
static int bpf_sk_storage_map_alloc_check(union bpf_attr *attr)
{
if (attr->map_flags & ~SK_STORAGE_CREATE_FLAG_MASK ||
@@ -619,12 +628,7 @@ static int bpf_sk_storage_map_alloc_check(union bpf_attr *attr)
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
- if (attr->value_size >= KMALLOC_MAX_SIZE -
- MAX_BPF_STACK - sizeof(struct bpf_sk_storage_elem) ||
- /* U16_MAX is much more than enough for sk local storage
- * considering a tcp_sock is ~2k.
- */
- attr->value_size > U16_MAX - sizeof(struct bpf_sk_storage_elem))
+ if (attr->value_size > MAX_VALUE_SIZE)
return -E2BIG;
return 0;
@@ -910,3 +914,270 @@ const struct bpf_func_proto bpf_sk_storage_delete_proto = {
.arg1_type = ARG_CONST_MAP_PTR,
.arg2_type = ARG_PTR_TO_SOCKET,
};
+
+struct bpf_sk_storage_diag {
+ u32 nr_maps;
+ struct bpf_map *maps[];
+};
+
+/* The reply will be like:
+ * INET_DIAG_BPF_SK_STORAGES (nla_nest)
+ * SK_DIAG_BPF_STORAGE (nla_nest)
+ * SK_DIAG_BPF_STORAGE_MAP_ID (nla_put_u32)
+ * SK_DIAG_BPF_STORAGE_MAP_VALUE (nla_reserve_64bit)
+ * SK_DIAG_BPF_STORAGE (nla_nest)
+ * SK_DIAG_BPF_STORAGE_MAP_ID (nla_put_u32)
+ * SK_DIAG_BPF_STORAGE_MAP_VALUE (nla_reserve_64bit)
+ * ....
+ */
+static int nla_value_size(u32 value_size)
+{
+ /* SK_DIAG_BPF_STORAGE (nla_nest)
+ * SK_DIAG_BPF_STORAGE_MAP_ID (nla_put_u32)
+ * SK_DIAG_BPF_STORAGE_MAP_VALUE (nla_reserve_64bit)
+ */
+ return nla_total_size(0) + nla_total_size(sizeof(u32)) +
+ nla_total_size_64bit(value_size);
+}
+
+void bpf_sk_storage_diag_free(struct bpf_sk_storage_diag *diag)
+{
+ u32 i;
+
+ if (!diag)
+ return;
+
+ for (i = 0; i < diag->nr_maps; i++)
+ bpf_map_put(diag->maps[i]);
+
+ kfree(diag);
+}
+EXPORT_SYMBOL_GPL(bpf_sk_storage_diag_free);
+
+static bool diag_check_dup(const struct bpf_sk_storage_diag *diag,
+ const struct bpf_map *map)
+{
+ u32 i;
+
+ for (i = 0; i < diag->nr_maps; i++) {
+ if (diag->maps[i] == map)
+ return true;
+ }
+
+ return false;
+}
+
+struct bpf_sk_storage_diag *
+bpf_sk_storage_diag_alloc(const struct nlattr *nla_stgs)
+{
+ struct bpf_sk_storage_diag *diag;
+ struct nlattr *nla;
+ u32 nr_maps = 0;
+ int rem, err;
+
+ /* bpf_sk_storage_map is currently limited to CAP_SYS_ADMIN as
+ * the map_alloc_check() side also does.
+ */
+ if (!capable(CAP_SYS_ADMIN))
+ return ERR_PTR(-EPERM);
+
+ nla_for_each_nested(nla, nla_stgs, rem) {
+ if (nla_type(nla) == SK_DIAG_BPF_STORAGE_REQ_MAP_FD)
+ nr_maps++;
+ }
+
+ diag = kzalloc(sizeof(*diag) + sizeof(diag->maps[0]) * nr_maps,
+ GFP_KERNEL);
+ if (!diag)
+ return ERR_PTR(-ENOMEM);
+
+ nla_for_each_nested(nla, nla_stgs, rem) {
+ struct bpf_map *map;
+ int map_fd;
+
+ if (nla_type(nla) != SK_DIAG_BPF_STORAGE_REQ_MAP_FD)
+ continue;
+
+ map_fd = nla_get_u32(nla);
+ map = bpf_map_get(map_fd);
+ if (IS_ERR(map)) {
+ err = PTR_ERR(map);
+ goto err_free;
+ }
+ if (map->map_type != BPF_MAP_TYPE_SK_STORAGE) {
+ bpf_map_put(map);
+ err = -EINVAL;
+ goto err_free;
+ }
+ if (diag_check_dup(diag, map)) {
+ bpf_map_put(map);
+ err = -EEXIST;
+ goto err_free;
+ }
+ diag->maps[diag->nr_maps++] = map;
+ }
+
+ return diag;
+
+err_free:
+ bpf_sk_storage_diag_free(diag);
+ return ERR_PTR(err);
+}
+EXPORT_SYMBOL_GPL(bpf_sk_storage_diag_alloc);
+
+static int diag_get(struct bpf_sk_storage_data *sdata, struct sk_buff *skb)
+{
+ struct nlattr *nla_stg, *nla_value;
+ struct bpf_sk_storage_map *smap;
+
+ /* It cannot exceed max nlattr's payload */
+ BUILD_BUG_ON(U16_MAX - NLA_HDRLEN < MAX_VALUE_SIZE);
+
+ nla_stg = nla_nest_start(skb, SK_DIAG_BPF_STORAGE);
+ if (!nla_stg)
+ return -EMSGSIZE;
+
+ smap = rcu_dereference(sdata->smap);
+ if (nla_put_u32(skb, SK_DIAG_BPF_STORAGE_MAP_ID, smap->map.id))
+ goto errout;
+
+ nla_value = nla_reserve_64bit(skb, SK_DIAG_BPF_STORAGE_MAP_VALUE,
+ smap->map.value_size,
+ SK_DIAG_BPF_STORAGE_PAD);
+ if (!nla_value)
+ goto errout;
+
+ if (map_value_has_spin_lock(&smap->map))
+ copy_map_value_locked(&smap->map, nla_data(nla_value),
+ sdata->data, true);
+ else
+ copy_map_value(&smap->map, nla_data(nla_value), sdata->data);
+
+ nla_nest_end(skb, nla_stg);
+ return 0;
+
+errout:
+ nla_nest_cancel(skb, nla_stg);
+ return -EMSGSIZE;
+}
+
+static int bpf_sk_storage_diag_put_all(struct sock *sk, struct sk_buff *skb,
+ int stg_array_type,
+ unsigned int *res_diag_size)
+{
+ /* stg_array_type (e.g. INET_DIAG_BPF_SK_STORAGES) */
+ unsigned int diag_size = nla_total_size(0);
+ struct bpf_sk_storage *sk_storage;
+ struct bpf_sk_storage_elem *selem;
+ struct bpf_sk_storage_map *smap;
+ struct nlattr *nla_stgs;
+ unsigned int saved_len;
+ int err = 0;
+
+ rcu_read_lock();
+
+ sk_storage = rcu_dereference(sk->sk_bpf_storage);
+ if (!sk_storage || hlist_empty(&sk_storage->list)) {
+ rcu_read_unlock();
+ return 0;
+ }
+
+ nla_stgs = nla_nest_start(skb, stg_array_type);
+ if (!nla_stgs)
+ /* Continue to learn diag_size */
+ err = -EMSGSIZE;
+
+ saved_len = skb->len;
+ hlist_for_each_entry_rcu(selem, &sk_storage->list, snode) {
+ smap = rcu_dereference(SDATA(selem)->smap);
+ diag_size += nla_value_size(smap->map.value_size);
+
+ if (nla_stgs && diag_get(SDATA(selem), skb))
+ /* Continue to learn diag_size */
+ err = -EMSGSIZE;
+ }
+
+ rcu_read_unlock();
+
+ if (nla_stgs) {
+ if (saved_len == skb->len)
+ nla_nest_cancel(skb, nla_stgs);
+ else
+ nla_nest_end(skb, nla_stgs);
+ }
+
+ if (diag_size == nla_total_size(0)) {
+ *res_diag_size = 0;
+ return 0;
+ }
+
+ *res_diag_size = diag_size;
+ return err;
+}
+
+int bpf_sk_storage_diag_put(struct bpf_sk_storage_diag *diag,
+ struct sock *sk, struct sk_buff *skb,
+ int stg_array_type,
+ unsigned int *res_diag_size)
+{
+ /* stg_array_type (e.g. INET_DIAG_BPF_SK_STORAGES) */
+ unsigned int diag_size = nla_total_size(0);
+ struct bpf_sk_storage *sk_storage;
+ struct bpf_sk_storage_data *sdata;
+ struct nlattr *nla_stgs;
+ unsigned int saved_len;
+ int err = 0;
+ u32 i;
+
+ *res_diag_size = 0;
+
+ /* No map has been specified. Dump all. */
+ if (!diag->nr_maps)
+ return bpf_sk_storage_diag_put_all(sk, skb, stg_array_type,
+ res_diag_size);
+
+ rcu_read_lock();
+ sk_storage = rcu_dereference(sk->sk_bpf_storage);
+ if (!sk_storage || hlist_empty(&sk_storage->list)) {
+ rcu_read_unlock();
+ return 0;
+ }
+
+ nla_stgs = nla_nest_start(skb, stg_array_type);
+ if (!nla_stgs)
+ /* Continue to learn diag_size */
+ err = -EMSGSIZE;
+
+ saved_len = skb->len;
+ for (i = 0; i < diag->nr_maps; i++) {
+ sdata = __sk_storage_lookup(sk_storage,
+ (struct bpf_sk_storage_map *)diag->maps[i],
+ false);
+
+ if (!sdata)
+ continue;
+
+ diag_size += nla_value_size(diag->maps[i]->value_size);
+
+ if (nla_stgs && diag_get(sdata, skb))
+ /* Continue to learn diag_size */
+ err = -EMSGSIZE;
+ }
+ rcu_read_unlock();
+
+ if (nla_stgs) {
+ if (saved_len == skb->len)
+ nla_nest_cancel(skb, nla_stgs);
+ else
+ nla_nest_end(skb, nla_stgs);
+ }
+
+ if (diag_size == nla_total_size(0)) {
+ *res_diag_size = 0;
+ return 0;
+ }
+
+ *res_diag_size = diag_size;
+ return err;
+}
+EXPORT_SYMBOL_GPL(bpf_sk_storage_diag_put);
diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
index a1670dff0629..3eff84824c8b 100644
--- a/net/core/flow_dissector.c
+++ b/net/core/flow_dissector.c
@@ -920,9 +920,7 @@ bool bpf_flow_dissect(struct bpf_prog *prog, struct bpf_flow_dissector *ctx,
(int)FLOW_DISSECTOR_F_STOP_AT_ENCAP);
flow_keys->flags = flags;
- preempt_disable();
- result = BPF_PROG_RUN(prog, ctx);
- preempt_enable();
+ result = bpf_prog_run_pin_on_cpu(prog, ctx);
flow_keys->nhoff = clamp_t(u16, flow_keys->nhoff, nhoff, hlen);
flow_keys->thoff = clamp_t(u16, flow_keys->thoff,
diff --git a/net/core/skmsg.c b/net/core/skmsg.c
index eeb28cb85664..c479372f2cd2 100644
--- a/net/core/skmsg.c
+++ b/net/core/skmsg.c
@@ -628,7 +628,6 @@ int sk_psock_msg_verdict(struct sock *sk, struct sk_psock *psock,
struct bpf_prog *prog;
int ret;
- preempt_disable();
rcu_read_lock();
prog = READ_ONCE(psock->progs.msg_parser);
if (unlikely(!prog)) {
@@ -638,7 +637,7 @@ int sk_psock_msg_verdict(struct sock *sk, struct sk_psock *psock,
sk_msg_compute_data_pointers(msg);
msg->sk = sk;
- ret = BPF_PROG_RUN(prog, msg);
+ ret = bpf_prog_run_pin_on_cpu(prog, msg);
ret = sk_psock_map_verd(ret, msg->sk_redir);
psock->apply_bytes = msg->apply_bytes;
if (ret == __SK_REDIRECT) {
@@ -653,7 +652,6 @@ int sk_psock_msg_verdict(struct sock *sk, struct sk_psock *psock,
}
out:
rcu_read_unlock();
- preempt_enable();
return ret;
}
EXPORT_SYMBOL_GPL(sk_psock_msg_verdict);
@@ -665,9 +663,7 @@ static int sk_psock_bpf_run(struct sk_psock *psock, struct bpf_prog *prog,
skb->sk = psock->sk;
bpf_compute_data_end_sk_skb(skb);
- preempt_disable();
- ret = BPF_PROG_RUN(prog, skb);
- preempt_enable();
+ ret = bpf_prog_run_pin_on_cpu(prog, skb);
/* strparser clones the skb before handing it to a upper layer,
* meaning skb_orphan has been called. We NULL sk on the way out
* to ensure we don't trigger a BUG_ON() in skb/sk operations
diff --git a/net/dccp/diag.c b/net/dccp/diag.c
index 73ef73a218ff..8a82c5a2c5a8 100644
--- a/net/dccp/diag.c
+++ b/net/dccp/diag.c
@@ -46,16 +46,15 @@ static void dccp_diag_get_info(struct sock *sk, struct inet_diag_msg *r,
}
static void dccp_diag_dump(struct sk_buff *skb, struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r, struct nlattr *bc)
+ const struct inet_diag_req_v2 *r)
{
- inet_diag_dump_icsk(&dccp_hashinfo, skb, cb, r, bc);
+ inet_diag_dump_icsk(&dccp_hashinfo, skb, cb, r);
}
-static int dccp_diag_dump_one(struct sk_buff *in_skb,
- const struct nlmsghdr *nlh,
+static int dccp_diag_dump_one(struct netlink_callback *cb,
const struct inet_diag_req_v2 *req)
{
- return inet_diag_dump_one_icsk(&dccp_hashinfo, in_skb, nlh, req);
+ return inet_diag_dump_one_icsk(&dccp_hashinfo, cb, req);
}
static const struct inet_diag_handler dccp_diag_handler = {
diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c
index f11e997e517b..e1cad25909df 100644
--- a/net/ipv4/inet_diag.c
+++ b/net/ipv4/inet_diag.c
@@ -23,6 +23,7 @@
#include <net/inet_hashtables.h>
#include <net/inet_timewait_sock.h>
#include <net/inet6_hashtables.h>
+#include <net/bpf_sk_storage.h>
#include <net/netlink.h>
#include <linux/inet.h>
@@ -156,26 +157,28 @@ errout:
}
EXPORT_SYMBOL_GPL(inet_diag_msg_attrs_fill);
+#define MAX_DUMP_ALLOC_SIZE (KMALLOC_MAX_SIZE - SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))
+
int inet_sk_diag_fill(struct sock *sk, struct inet_connection_sock *icsk,
- struct sk_buff *skb, const struct inet_diag_req_v2 *req,
- struct user_namespace *user_ns,
- u32 portid, u32 seq, u16 nlmsg_flags,
- const struct nlmsghdr *unlh,
- bool net_admin)
+ struct sk_buff *skb, struct netlink_callback *cb,
+ const struct inet_diag_req_v2 *req,
+ u16 nlmsg_flags, bool net_admin)
{
const struct tcp_congestion_ops *ca_ops;
const struct inet_diag_handler *handler;
+ struct inet_diag_dump_data *cb_data;
int ext = req->idiag_ext;
struct inet_diag_msg *r;
struct nlmsghdr *nlh;
struct nlattr *attr;
void *info = NULL;
+ cb_data = cb->data;
handler = inet_diag_table[req->sdiag_protocol];
BUG_ON(!handler);
- nlh = nlmsg_put(skb, portid, seq, unlh->nlmsg_type, sizeof(*r),
- nlmsg_flags);
+ nlh = nlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
+ cb->nlh->nlmsg_type, sizeof(*r), nlmsg_flags);
if (!nlh)
return -EMSGSIZE;
@@ -187,7 +190,9 @@ int inet_sk_diag_fill(struct sock *sk, struct inet_connection_sock *icsk,
r->idiag_timer = 0;
r->idiag_retrans = 0;
- if (inet_diag_msg_attrs_fill(sk, skb, r, ext, user_ns, net_admin))
+ if (inet_diag_msg_attrs_fill(sk, skb, r, ext,
+ sk_user_ns(NETLINK_CB(cb->skb).sk),
+ net_admin))
goto errout;
if (ext & (1 << (INET_DIAG_MEMINFO - 1))) {
@@ -302,6 +307,48 @@ int inet_sk_diag_fill(struct sock *sk, struct inet_connection_sock *icsk,
goto errout;
}
+ /* Keep it at the end for potential retry with a larger skb,
+ * or else do best-effort fitting, which is only done for the
+ * first_nlmsg.
+ */
+ if (cb_data->bpf_stg_diag) {
+ bool first_nlmsg = ((unsigned char *)nlh == skb->data);
+ unsigned int prev_min_dump_alloc;
+ unsigned int total_nla_size = 0;
+ unsigned int msg_len;
+ int err;
+
+ msg_len = skb_tail_pointer(skb) - (unsigned char *)nlh;
+ err = bpf_sk_storage_diag_put(cb_data->bpf_stg_diag, sk, skb,
+ INET_DIAG_SK_BPF_STORAGES,
+ &total_nla_size);
+
+ if (!err)
+ goto out;
+
+ total_nla_size += msg_len;
+ prev_min_dump_alloc = cb->min_dump_alloc;
+ if (total_nla_size > prev_min_dump_alloc)
+ cb->min_dump_alloc = min_t(u32, total_nla_size,
+ MAX_DUMP_ALLOC_SIZE);
+
+ if (!first_nlmsg)
+ goto errout;
+
+ if (cb->min_dump_alloc > prev_min_dump_alloc)
+ /* Retry with pskb_expand_head() with
+ * __GFP_DIRECT_RECLAIM
+ */
+ goto errout;
+
+ WARN_ON_ONCE(total_nla_size <= prev_min_dump_alloc);
+
+ /* Send what we have for this sk
+ * and move on to the next sk in the following
+ * dump()
+ */
+ }
+
out:
nlmsg_end(skb, nlh);
return 0;
@@ -312,30 +359,19 @@ errout:
}
EXPORT_SYMBOL_GPL(inet_sk_diag_fill);
-static int inet_csk_diag_fill(struct sock *sk,
- struct sk_buff *skb,
- const struct inet_diag_req_v2 *req,
- struct user_namespace *user_ns,
- u32 portid, u32 seq, u16 nlmsg_flags,
- const struct nlmsghdr *unlh,
- bool net_admin)
-{
- return inet_sk_diag_fill(sk, inet_csk(sk), skb, req, user_ns,
- portid, seq, nlmsg_flags, unlh, net_admin);
-}
-
static int inet_twsk_diag_fill(struct sock *sk,
struct sk_buff *skb,
- u32 portid, u32 seq, u16 nlmsg_flags,
- const struct nlmsghdr *unlh)
+ struct netlink_callback *cb,
+ u16 nlmsg_flags)
{
struct inet_timewait_sock *tw = inet_twsk(sk);
struct inet_diag_msg *r;
struct nlmsghdr *nlh;
long tmo;
- nlh = nlmsg_put(skb, portid, seq, unlh->nlmsg_type, sizeof(*r),
- nlmsg_flags);
+ nlh = nlmsg_put(skb, NETLINK_CB(cb->skb).portid,
+ cb->nlh->nlmsg_seq, cb->nlh->nlmsg_type,
+ sizeof(*r), nlmsg_flags);
if (!nlh)
return -EMSGSIZE;
@@ -359,16 +395,16 @@ static int inet_twsk_diag_fill(struct sock *sk,
}
static int inet_req_diag_fill(struct sock *sk, struct sk_buff *skb,
- u32 portid, u32 seq, u16 nlmsg_flags,
- const struct nlmsghdr *unlh, bool net_admin)
+ struct netlink_callback *cb,
+ u16 nlmsg_flags, bool net_admin)
{
struct request_sock *reqsk = inet_reqsk(sk);
struct inet_diag_msg *r;
struct nlmsghdr *nlh;
long tmo;
- nlh = nlmsg_put(skb, portid, seq, unlh->nlmsg_type, sizeof(*r),
- nlmsg_flags);
+ nlh = nlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
+ cb->nlh->nlmsg_type, sizeof(*r), nlmsg_flags);
if (!nlh)
return -EMSGSIZE;
@@ -397,21 +433,18 @@ static int inet_req_diag_fill(struct sock *sk, struct sk_buff *skb,
}
static int sk_diag_fill(struct sock *sk, struct sk_buff *skb,
+ struct netlink_callback *cb,
const struct inet_diag_req_v2 *r,
- struct user_namespace *user_ns,
- u32 portid, u32 seq, u16 nlmsg_flags,
- const struct nlmsghdr *unlh, bool net_admin)
+ u16 nlmsg_flags, bool net_admin)
{
if (sk->sk_state == TCP_TIME_WAIT)
- return inet_twsk_diag_fill(sk, skb, portid, seq,
- nlmsg_flags, unlh);
+ return inet_twsk_diag_fill(sk, skb, cb, nlmsg_flags);
if (sk->sk_state == TCP_NEW_SYN_RECV)
- return inet_req_diag_fill(sk, skb, portid, seq,
- nlmsg_flags, unlh, net_admin);
+ return inet_req_diag_fill(sk, skb, cb, nlmsg_flags, net_admin);
- return inet_csk_diag_fill(sk, skb, r, user_ns, portid, seq,
- nlmsg_flags, unlh, net_admin);
+ return inet_sk_diag_fill(sk, inet_csk(sk), skb, cb, r, nlmsg_flags,
+ net_admin);
}
struct sock *inet_diag_find_one_icsk(struct net *net,
@@ -459,10 +492,10 @@ struct sock *inet_diag_find_one_icsk(struct net *net,
EXPORT_SYMBOL_GPL(inet_diag_find_one_icsk);
int inet_diag_dump_one_icsk(struct inet_hashinfo *hashinfo,
- struct sk_buff *in_skb,
- const struct nlmsghdr *nlh,
+ struct netlink_callback *cb,
const struct inet_diag_req_v2 *req)
{
+ struct sk_buff *in_skb = cb->skb;
bool net_admin = netlink_net_capable(in_skb, CAP_NET_ADMIN);
struct net *net = sock_net(in_skb->sk);
struct sk_buff *rep;
@@ -479,10 +512,7 @@ int inet_diag_dump_one_icsk(struct inet_hashinfo *hashinfo,
goto out;
}
- err = sk_diag_fill(sk, rep, req,
- sk_user_ns(NETLINK_CB(in_skb).sk),
- NETLINK_CB(in_skb).portid,
- nlh->nlmsg_seq, 0, nlh, net_admin);
+ err = sk_diag_fill(sk, rep, cb, req, 0, net_admin);
if (err < 0) {
WARN_ON(err == -EMSGSIZE);
nlmsg_free(rep);
@@ -509,14 +539,21 @@ static int inet_diag_cmd_exact(int cmd, struct sk_buff *in_skb,
int err;
handler = inet_diag_lock_handler(req->sdiag_protocol);
- if (IS_ERR(handler))
+ if (IS_ERR(handler)) {
err = PTR_ERR(handler);
- else if (cmd == SOCK_DIAG_BY_FAMILY)
- err = handler->dump_one(in_skb, nlh, req);
- else if (cmd == SOCK_DESTROY && handler->destroy)
+ } else if (cmd == SOCK_DIAG_BY_FAMILY) {
+ struct inet_diag_dump_data empty_dump_data = {};
+ struct netlink_callback cb = {
+ .nlh = nlh,
+ .skb = in_skb,
+ .data = &empty_dump_data,
+ };
+ err = handler->dump_one(&cb, req);
+ } else if (cmd == SOCK_DESTROY && handler->destroy) {
err = handler->destroy(in_skb, req);
- else
+ } else {
err = -EOPNOTSUPP;
+ }
inet_diag_unlock_handler(handler);
return err;
@@ -847,23 +884,6 @@ static int inet_diag_bc_audit(const struct nlattr *attr,
return len == 0 ? 0 : -EINVAL;
}
-static int inet_csk_diag_dump(struct sock *sk,
- struct sk_buff *skb,
- struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r,
- const struct nlattr *bc,
- bool net_admin)
-{
- if (!inet_diag_bc_sk(bc, sk))
- return 0;
-
- return inet_csk_diag_fill(sk, skb, r,
- sk_user_ns(NETLINK_CB(cb->skb).sk),
- NETLINK_CB(cb->skb).portid,
- cb->nlh->nlmsg_seq, NLM_F_MULTI, cb->nlh,
- net_admin);
-}
-
static void twsk_build_assert(void)
{
BUILD_BUG_ON(offsetof(struct inet_timewait_sock, tw_family) !=
@@ -892,14 +912,17 @@ static void twsk_build_assert(void)
void inet_diag_dump_icsk(struct inet_hashinfo *hashinfo, struct sk_buff *skb,
struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r, struct nlattr *bc)
+ const struct inet_diag_req_v2 *r)
{
bool net_admin = netlink_net_capable(cb->skb, CAP_NET_ADMIN);
+ struct inet_diag_dump_data *cb_data = cb->data;
struct net *net = sock_net(skb->sk);
u32 idiag_states = r->idiag_states;
int i, num, s_i, s_num;
+ struct nlattr *bc;
struct sock *sk;
+ bc = cb_data->inet_diag_nla_bc;
if (idiag_states & TCPF_SYN_RECV)
idiag_states |= TCPF_NEW_SYN_RECV;
s_i = cb->args[1];
@@ -935,8 +958,12 @@ void inet_diag_dump_icsk(struct inet_hashinfo *hashinfo, struct sk_buff *skb,
r->id.idiag_sport)
goto next_listen;
- if (inet_csk_diag_dump(sk, skb, cb, r,
- bc, net_admin) < 0) {
+ if (!inet_diag_bc_sk(bc, sk))
+ goto next_listen;
+
+ if (inet_sk_diag_fill(sk, inet_csk(sk), skb,
+ cb, r, NLM_F_MULTI,
+ net_admin) < 0) {
spin_unlock(&ilb->lock);
goto done;
}
@@ -1014,11 +1041,8 @@ next_normal:
res = 0;
for (idx = 0; idx < accum; idx++) {
if (res >= 0) {
- res = sk_diag_fill(sk_arr[idx], skb, r,
- sk_user_ns(NETLINK_CB(cb->skb).sk),
- NETLINK_CB(cb->skb).portid,
- cb->nlh->nlmsg_seq, NLM_F_MULTI,
- cb->nlh, net_admin);
+ res = sk_diag_fill(sk_arr[idx], skb, cb, r,
+ NLM_F_MULTI, net_admin);
if (res < 0)
num = num_arr[idx];
}
@@ -1042,31 +1066,101 @@ out:
EXPORT_SYMBOL_GPL(inet_diag_dump_icsk);
static int __inet_diag_dump(struct sk_buff *skb, struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r,
- struct nlattr *bc)
+ const struct inet_diag_req_v2 *r)
{
const struct inet_diag_handler *handler;
+ u32 prev_min_dump_alloc;
int err = 0;
+again:
+ prev_min_dump_alloc = cb->min_dump_alloc;
handler = inet_diag_lock_handler(r->sdiag_protocol);
if (!IS_ERR(handler))
- handler->dump(skb, cb, r, bc);
+ handler->dump(skb, cb, r);
else
err = PTR_ERR(handler);
inet_diag_unlock_handler(handler);
+ /* The skb is not large enough to fit one sk info and
+ * inet_sk_diag_fill() has requested for a larger skb.
+ */
+ if (!skb->len && cb->min_dump_alloc > prev_min_dump_alloc) {
+ err = pskb_expand_head(skb, 0, cb->min_dump_alloc, GFP_KERNEL);
+ if (!err)
+ goto again;
+ }
+
return err ? : skb->len;
}
static int inet_diag_dump(struct sk_buff *skb, struct netlink_callback *cb)
{
- int hdrlen = sizeof(struct inet_diag_req_v2);
- struct nlattr *bc = NULL;
+ return __inet_diag_dump(skb, cb, nlmsg_data(cb->nlh));
+}
- if (nlmsg_attrlen(cb->nlh, hdrlen))
- bc = nlmsg_find_attr(cb->nlh, hdrlen, INET_DIAG_REQ_BYTECODE);
+static int __inet_diag_dump_start(struct netlink_callback *cb, int hdrlen)
+{
+ const struct nlmsghdr *nlh = cb->nlh;
+ struct inet_diag_dump_data *cb_data;
+ struct sk_buff *skb = cb->skb;
+ struct nlattr *nla;
+ int rem, err;
+
+ cb_data = kzalloc(sizeof(*cb_data), GFP_KERNEL);
+ if (!cb_data)
+ return -ENOMEM;
+
+ nla_for_each_attr(nla, nlmsg_attrdata(nlh, hdrlen),
+ nlmsg_attrlen(nlh, hdrlen), rem) {
+ int type = nla_type(nla);
+
+ if (type < __INET_DIAG_REQ_MAX)
+ cb_data->req_nlas[type] = nla;
+ }
+
+ nla = cb_data->inet_diag_nla_bc;
+ if (nla) {
+ err = inet_diag_bc_audit(nla, skb);
+ if (err) {
+ kfree(cb_data);
+ return err;
+ }
+ }
+
+ nla = cb_data->inet_diag_nla_bpf_stgs;
+ if (nla) {
+ struct bpf_sk_storage_diag *bpf_stg_diag;
+
+ bpf_stg_diag = bpf_sk_storage_diag_alloc(nla);
+ if (IS_ERR(bpf_stg_diag)) {
+ kfree(cb_data);
+ return PTR_ERR(bpf_stg_diag);
+ }
+ cb_data->bpf_stg_diag = bpf_stg_diag;
+ }
+
+ cb->data = cb_data;
+ return 0;
+}
+
+static int inet_diag_dump_start(struct netlink_callback *cb)
+{
+ return __inet_diag_dump_start(cb, sizeof(struct inet_diag_req_v2));
+}
+
+static int inet_diag_dump_start_compat(struct netlink_callback *cb)
+{
+ return __inet_diag_dump_start(cb, sizeof(struct inet_diag_req));
+}
- return __inet_diag_dump(skb, cb, nlmsg_data(cb->nlh), bc);
+static int inet_diag_dump_done(struct netlink_callback *cb)
+{
+ struct inet_diag_dump_data *cb_data = cb->data;
+
+ bpf_sk_storage_diag_free(cb_data->bpf_stg_diag);
+ kfree(cb->data);
+
+ return 0;
}
static int inet_diag_type2proto(int type)
@@ -1085,9 +1179,7 @@ static int inet_diag_dump_compat(struct sk_buff *skb,
struct netlink_callback *cb)
{
struct inet_diag_req *rc = nlmsg_data(cb->nlh);
- int hdrlen = sizeof(struct inet_diag_req);
struct inet_diag_req_v2 req;
- struct nlattr *bc = NULL;
req.sdiag_family = AF_UNSPEC; /* compatibility */
req.sdiag_protocol = inet_diag_type2proto(cb->nlh->nlmsg_type);
@@ -1095,10 +1187,7 @@ static int inet_diag_dump_compat(struct sk_buff *skb,
req.idiag_states = rc->idiag_states;
req.id = rc->id;
- if (nlmsg_attrlen(cb->nlh, hdrlen))
- bc = nlmsg_find_attr(cb->nlh, hdrlen, INET_DIAG_REQ_BYTECODE);
-
- return __inet_diag_dump(skb, cb, &req, bc);
+ return __inet_diag_dump(skb, cb, &req);
}
static int inet_diag_get_exact_compat(struct sk_buff *in_skb,
@@ -1126,22 +1215,12 @@ static int inet_diag_rcv_msg_compat(struct sk_buff *skb, struct nlmsghdr *nlh)
return -EINVAL;
if (nlh->nlmsg_flags & NLM_F_DUMP) {
- if (nlmsg_attrlen(nlh, hdrlen)) {
- struct nlattr *attr;
- int err;
-
- attr = nlmsg_find_attr(nlh, hdrlen,
- INET_DIAG_REQ_BYTECODE);
- err = inet_diag_bc_audit(attr, skb);
- if (err)
- return err;
- }
- {
- struct netlink_dump_control c = {
- .dump = inet_diag_dump_compat,
- };
- return netlink_dump_start(net->diag_nlsk, skb, nlh, &c);
- }
+ struct netlink_dump_control c = {
+ .start = inet_diag_dump_start_compat,
+ .done = inet_diag_dump_done,
+ .dump = inet_diag_dump_compat,
+ };
+ return netlink_dump_start(net->diag_nlsk, skb, nlh, &c);
}
return inet_diag_get_exact_compat(skb, nlh);
@@ -1157,22 +1236,12 @@ static int inet_diag_handler_cmd(struct sk_buff *skb, struct nlmsghdr *h)
if (h->nlmsg_type == SOCK_DIAG_BY_FAMILY &&
h->nlmsg_flags & NLM_F_DUMP) {
- if (nlmsg_attrlen(h, hdrlen)) {
- struct nlattr *attr;
- int err;
-
- attr = nlmsg_find_attr(h, hdrlen,
- INET_DIAG_REQ_BYTECODE);
- err = inet_diag_bc_audit(attr, skb);
- if (err)
- return err;
- }
- {
- struct netlink_dump_control c = {
- .dump = inet_diag_dump,
- };
- return netlink_dump_start(net->diag_nlsk, skb, h, &c);
- }
+ struct netlink_dump_control c = {
+ .start = inet_diag_dump_start,
+ .done = inet_diag_dump_done,
+ .dump = inet_diag_dump,
+ };
+ return netlink_dump_start(net->diag_nlsk, skb, h, &c);
}
return inet_diag_cmd_exact(h->nlmsg_type, skb, h, nlmsg_data(h));
diff --git a/net/ipv4/raw_diag.c b/net/ipv4/raw_diag.c
index e35736b99300..d19cce39be1b 100644
--- a/net/ipv4/raw_diag.c
+++ b/net/ipv4/raw_diag.c
@@ -87,15 +87,16 @@ out_unlock:
return sk ? sk : ERR_PTR(-ENOENT);
}
-static int raw_diag_dump_one(struct sk_buff *in_skb,
- const struct nlmsghdr *nlh,
+static int raw_diag_dump_one(struct netlink_callback *cb,
const struct inet_diag_req_v2 *r)
{
- struct net *net = sock_net(in_skb->sk);
+ struct sk_buff *in_skb = cb->skb;
struct sk_buff *rep;
struct sock *sk;
+ struct net *net;
int err;
+ net = sock_net(in_skb->sk);
sk = raw_sock_get(net, r);
if (IS_ERR(sk))
return PTR_ERR(sk);
@@ -108,10 +109,7 @@ static int raw_diag_dump_one(struct sk_buff *in_skb,
return -ENOMEM;
}
- err = inet_sk_diag_fill(sk, NULL, rep, r,
- sk_user_ns(NETLINK_CB(in_skb).sk),
- NETLINK_CB(in_skb).portid,
- nlh->nlmsg_seq, 0, nlh,
+ err = inet_sk_diag_fill(sk, NULL, rep, cb, r, 0,
netlink_net_capable(in_skb, CAP_NET_ADMIN));
sock_put(sk);
@@ -136,25 +134,25 @@ static int sk_diag_dump(struct sock *sk, struct sk_buff *skb,
if (!inet_diag_bc_sk(bc, sk))
return 0;
- return inet_sk_diag_fill(sk, NULL, skb, r,
- sk_user_ns(NETLINK_CB(cb->skb).sk),
- NETLINK_CB(cb->skb).portid,
- cb->nlh->nlmsg_seq, NLM_F_MULTI,
- cb->nlh, net_admin);
+ return inet_sk_diag_fill(sk, NULL, skb, cb, r, NLM_F_MULTI, net_admin);
}
static void raw_diag_dump(struct sk_buff *skb, struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r, struct nlattr *bc)
+ const struct inet_diag_req_v2 *r)
{
bool net_admin = netlink_net_capable(cb->skb, CAP_NET_ADMIN);
struct raw_hashinfo *hashinfo = raw_get_hashinfo(r);
struct net *net = sock_net(skb->sk);
+ struct inet_diag_dump_data *cb_data;
int num, s_num, slot, s_slot;
struct sock *sk = NULL;
+ struct nlattr *bc;
if (IS_ERR(hashinfo))
return;
+ cb_data = cb->data;
+ bc = cb_data->inet_diag_nla_bc;
s_slot = cb->args[0];
num = s_num = cb->args[1];
diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c
index 0d08f9e2d8d0..75a1c985f49a 100644
--- a/net/ipv4/tcp_diag.c
+++ b/net/ipv4/tcp_diag.c
@@ -179,15 +179,15 @@ static size_t tcp_diag_get_aux_size(struct sock *sk, bool net_admin)
}
static void tcp_diag_dump(struct sk_buff *skb, struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r, struct nlattr *bc)
+ const struct inet_diag_req_v2 *r)
{
- inet_diag_dump_icsk(&tcp_hashinfo, skb, cb, r, bc);
+ inet_diag_dump_icsk(&tcp_hashinfo, skb, cb, r);
}
-static int tcp_diag_dump_one(struct sk_buff *in_skb, const struct nlmsghdr *nlh,
+static int tcp_diag_dump_one(struct netlink_callback *cb,
const struct inet_diag_req_v2 *req)
{
- return inet_diag_dump_one_icsk(&tcp_hashinfo, in_skb, nlh, req);
+ return inet_diag_dump_one_icsk(&tcp_hashinfo, cb, req);
}
#ifdef CONFIG_INET_DIAG_DESTROY
diff --git a/net/ipv4/udp_diag.c b/net/ipv4/udp_diag.c
index 910555a4d9fe..93884696abdd 100644
--- a/net/ipv4/udp_diag.c
+++ b/net/ipv4/udp_diag.c
@@ -21,16 +21,15 @@ static int sk_diag_dump(struct sock *sk, struct sk_buff *skb,
if (!inet_diag_bc_sk(bc, sk))
return 0;
- return inet_sk_diag_fill(sk, NULL, skb, req,
- sk_user_ns(NETLINK_CB(cb->skb).sk),
- NETLINK_CB(cb->skb).portid,
- cb->nlh->nlmsg_seq, NLM_F_MULTI, cb->nlh, net_admin);
+ return inet_sk_diag_fill(sk, NULL, skb, cb, req, NLM_F_MULTI,
+ net_admin);
}
-static int udp_dump_one(struct udp_table *tbl, struct sk_buff *in_skb,
- const struct nlmsghdr *nlh,
+static int udp_dump_one(struct udp_table *tbl,
+ struct netlink_callback *cb,
const struct inet_diag_req_v2 *req)
{
+ struct sk_buff *in_skb = cb->skb;
int err = -EINVAL;
struct sock *sk = NULL;
struct sk_buff *rep;
@@ -70,11 +69,8 @@ static int udp_dump_one(struct udp_table *tbl, struct sk_buff *in_skb,
if (!rep)
goto out;
- err = inet_sk_diag_fill(sk, NULL, rep, req,
- sk_user_ns(NETLINK_CB(in_skb).sk),
- NETLINK_CB(in_skb).portid,
- nlh->nlmsg_seq, 0, nlh,
- netlink_net_capable(in_skb, CAP_NET_ADMIN));
+ err = inet_sk_diag_fill(sk, NULL, rep, cb, req, 0,
+ netlink_net_capable(in_skb, CAP_NET_ADMIN));
if (err < 0) {
WARN_ON(err == -EMSGSIZE);
kfree_skb(rep);
@@ -93,12 +89,16 @@ out_nosk:
static void udp_dump(struct udp_table *table, struct sk_buff *skb,
struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r, struct nlattr *bc)
+ const struct inet_diag_req_v2 *r)
{
bool net_admin = netlink_net_capable(cb->skb, CAP_NET_ADMIN);
struct net *net = sock_net(skb->sk);
+ struct inet_diag_dump_data *cb_data;
int num, s_num, slot, s_slot;
+ struct nlattr *bc;
+ cb_data = cb->data;
+ bc = cb_data->inet_diag_nla_bc;
s_slot = cb->args[0];
num = s_num = cb->args[1];
@@ -146,15 +146,15 @@ done:
}
static void udp_diag_dump(struct sk_buff *skb, struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r, struct nlattr *bc)
+ const struct inet_diag_req_v2 *r)
{
- udp_dump(&udp_table, skb, cb, r, bc);
+ udp_dump(&udp_table, skb, cb, r);
}
-static int udp_diag_dump_one(struct sk_buff *in_skb, const struct nlmsghdr *nlh,
+static int udp_diag_dump_one(struct netlink_callback *cb,
const struct inet_diag_req_v2 *req)
{
- return udp_dump_one(&udp_table, in_skb, nlh, req);
+ return udp_dump_one(&udp_table, cb, req);
}
static void udp_diag_get_info(struct sock *sk, struct inet_diag_msg *r,
@@ -249,16 +249,15 @@ static const struct inet_diag_handler udp_diag_handler = {
};
static void udplite_diag_dump(struct sk_buff *skb, struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r,
- struct nlattr *bc)
+ const struct inet_diag_req_v2 *r)
{
- udp_dump(&udplite_table, skb, cb, r, bc);
+ udp_dump(&udplite_table, skb, cb, r);
}
-static int udplite_diag_dump_one(struct sk_buff *in_skb, const struct nlmsghdr *nlh,
+static int udplite_diag_dump_one(struct netlink_callback *cb,
const struct inet_diag_req_v2 *req)
{
- return udp_dump_one(&udplite_table, in_skb, nlh, req);
+ return udp_dump_one(&udplite_table, cb, req);
}
static const struct inet_diag_handler udplite_diag_handler = {
diff --git a/net/kcm/kcmsock.c b/net/kcm/kcmsock.c
index ea9e73428ed9..56fac24a627a 100644
--- a/net/kcm/kcmsock.c
+++ b/net/kcm/kcmsock.c
@@ -380,9 +380,7 @@ static int kcm_parse_func_strparser(struct strparser *strp, struct sk_buff *skb)
struct bpf_prog *prog = psock->bpf_prog;
int res;
- preempt_disable();
- res = BPF_PROG_RUN(prog, skb);
- preempt_enable();
+ res = bpf_prog_run_pin_on_cpu(prog, skb);
return res;
}
diff --git a/net/sctp/diag.c b/net/sctp/diag.c
index 8a15146faaeb..69743a6aaf6f 100644
--- a/net/sctp/diag.c
+++ b/net/sctp/diag.c
@@ -432,11 +432,12 @@ static void sctp_diag_get_info(struct sock *sk, struct inet_diag_msg *r,
sctp_get_sctp_info(sk, infox->asoc, infox->sctpinfo);
}
-static int sctp_diag_dump_one(struct sk_buff *in_skb,
- const struct nlmsghdr *nlh,
+static int sctp_diag_dump_one(struct netlink_callback *cb,
const struct inet_diag_req_v2 *req)
{
+ struct sk_buff *in_skb = cb->skb;
struct net *net = sock_net(in_skb->sk);
+ const struct nlmsghdr *nlh = cb->nlh;
union sctp_addr laddr, paddr;
struct sctp_comm_param commp = {
.skb = in_skb,
@@ -470,7 +471,7 @@ static int sctp_diag_dump_one(struct sk_buff *in_skb,
}
static void sctp_diag_dump(struct sk_buff *skb, struct netlink_callback *cb,
- const struct inet_diag_req_v2 *r, struct nlattr *bc)
+ const struct inet_diag_req_v2 *r)
{
u32 idiag_states = r->idiag_states;
struct net *net = sock_net(skb->sk);