* Re: [PATCH] igb: Fix format with line continuation whitespace
From: Vinicius Costa Gomes @ 2018-11-15 1:04 UTC (permalink / raw)
To: Joe Perches, Jeff Kirsher
Cc: David S. Miller, intel-wired-lan, netdev, linux-kernel
In-Reply-To: <1b31c663afa1aa7b3d231917f00f58b9cb604231.camel@perches.com>
Joe Perches <joe@perches.com> writes:
> The line continuation unintentionally adds whitespace so
> instead use a coalesced format to remove the whitespace.
>
> Miscellanea:
>
> o Use a more typical style for ternaries and arguments
> for this logging message
>
> Signed-off-by: Joe Perches <joe@perches.com>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Cheers,
--
Vinicius
^ permalink raw reply
* Re: [PATCH][bpf-next] bpf: fix null pointer dereference on pointer offload
From: Jakub Kicinski @ 2018-11-15 0:52 UTC (permalink / raw)
To: Colin King
Cc: Alexei Starovoitov, Daniel Borkmann, netdev, kernel-janitors,
linux-kernel
In-Reply-To: <20181113092926.27081-1-colin.king@canonical.com>
On Tue, 13 Nov 2018 09:29:26 +0000, Colin King wrote:
> From: Colin Ian King <colin.king@canonical.com>
>
> Pointer offload is being null checked however the following statement
> dereferences the potentially null pointer offload when assigning
> offload->dev_state. Fix this by only assigning it if offload is not
> null.
>
> Detected by CoverityScan, CID#1475437 ("Dereference after null check")
>
> Fixes: 00db12c3d141 ("bpf: call verifier_prep from its callback in struct bpf_offload_dev")
> Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Thanks!
^ permalink raw reply
* [PATCH net] qed: Fix qed compilation issue when CONFIG_QED_RDMA not defined
From: Denis Bolotin @ 2018-11-14 14:36 UTC (permalink / raw)
To: davem, netdev; +Cc: ariel.elior, Denis Bolotin
Add a missing semicolon to a line in an empty implementation function.
Signed-off-by: Denis Bolotin <denis.bolotin@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
---
drivers/net/ethernet/qlogic/qed/qed_rdma.h | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/qlogic/qed/qed_rdma.h b/drivers/net/ethernet/qlogic/qed/qed_rdma.h
index 50d609c..5eec88c 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_rdma.h
+++ b/drivers/net/ethernet/qlogic/qed/qed_rdma.h
@@ -183,7 +183,10 @@ struct qed_rdma_qp {
static inline void qed_rdma_dpm_conf(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt) {}
static inline void qed_rdma_dpm_bar(struct qed_hwfn *p_hwfn,
struct qed_ptt *p_ptt) {}
-static inline int qed_rdma_info_alloc(struct qed_hwfn *p_hwfn) {return -EINVAL}
+static inline int qed_rdma_info_alloc(struct qed_hwfn *p_hwfn)
+{
+ return -EINVAL;
+}
static inline void qed_rdma_info_free(struct qed_hwfn *p_hwfn) {}
#endif
--
1.8.3.1
^ permalink raw reply related
* [PATCH] allow DSCP values in ip rulesB
From: Pavel Balaev @ 2018-11-14 14:30 UTC (permalink / raw)
To: netdev
Hello, for now IP rules supports only old TOS values and we cannot use
DSCP.
This patch adds support for DSCP values in IP rules:
$ ip r add default via 192.168.0.6 table test
$ ip ru add tos 0x80 table test
$ ip ru
0: from all lookup local
32764: from all tos CS4 lookup test
32766: from all lookup main
32767: from all lookup default
$ ip r get fibmatch 8.8.8.9 tos 0x80
default tos CS4 via 192.168.0.6 dev lan table test
Signed-off-by: Pavel Balaev <mail@void.so>
---
net/ipv4/fib_rules.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index f8eb78d0..7a6c5bfe 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -220,7 +220,7 @@ static int fib4_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
int err = -EINVAL;
struct fib4_rule *rule4 = (struct fib4_rule *) rule;
- if (frh->tos & ~IPTOS_TOS_MASK) {
+ if (frh->tos & ~(IPTOS_TOS_MASK | IPTOS_PREC_MASK)) {
NL_SET_ERR_MSG(extack, "Invalid tos");
goto errout;
}
--
2.18.1
^ permalink raw reply related
* RE: [PATCH] igb: Fix format with line continuation whitespace
From: Brown, Aaron F @ 2018-11-15 0:17 UTC (permalink / raw)
To: Joe Perches, Kirsher, Jeffrey T
Cc: David S. Miller, intel-wired-lan@lists.osuosl.org,
netdev@vger.kernel.org, linux-kernel@vger.kernel.org
In-Reply-To: <1b31c663afa1aa7b3d231917f00f58b9cb604231.camel@perches.com>
> From: netdev-owner@vger.kernel.org [mailto:netdev-
> owner@vger.kernel.org] On Behalf Of Joe Perches
> Sent: Thursday, November 1, 2018 12:04 AM
> To: Kirsher, Jeffrey T <jeffrey.t.kirsher@intel.com>
> Cc: David S. Miller <davem@davemloft.net>; intel-wired-
> lan@lists.osuosl.org; netdev@vger.kernel.org; linux-kernel@vger.kernel.org
> Subject: [PATCH] igb: Fix format with line continuation whitespace
>
> The line continuation unintentionally adds whitespace so
> instead use a coalesced format to remove the whitespace.
>
> Miscellanea:
>
> o Use a more typical style for ternaries and arguments
> for this logging message
>
> Signed-off-by: Joe Perches <joe@perches.com>
> ---
> drivers/net/ethernet/intel/igb/igb_main.c | 13 ++++++-------
> 1 file changed, 6 insertions(+), 7 deletions(-)
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
^ permalink raw reply
* [PATCH 16/16] batman-adv: enable MCAST by default at compile time
From: Simon Wunderlich @ 2018-11-14 14:08 UTC (permalink / raw)
To: davem
Cc: netdev, b.a.t.m.a.n, Linus Lüssing, Sven Eckelmann,
Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Linus Lüssing <linus.luessing@c0d3.blue>
Thanks to rigorous testing in wireless community mesh networks several
issues with multicast entries in the translation table were found and
fixed in the last 1.5 years. Now we see the first larger networks
(a few hundred nodes) with a batman-adv version with multicast
optimizations enabled arising, with no TT / multicast optimization
related issues so far.
Therefore it seems safe to enable multicast optimizations by default.
Signed-off-by: Linus Lüssing <linus.luessing@c0d3.blue>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/Kconfig | 1 +
1 file changed, 1 insertion(+)
diff --git a/net/batman-adv/Kconfig b/net/batman-adv/Kconfig
index d6b94559f888..c386e6981416 100644
--- a/net/batman-adv/Kconfig
+++ b/net/batman-adv/Kconfig
@@ -82,6 +82,7 @@ config BATMAN_ADV_NC
config BATMAN_ADV_MCAST
bool "Multicast optimisation"
depends on BATMAN_ADV && INET && !(BRIDGE=m && BATMAN_ADV=y)
+ default y
help
This option enables the multicast optimisation which aims to
reduce the air overhead while improving the reliability of
--
2.11.0
^ permalink raw reply related
* [PATCH 15/16] batman-adv: Move CRC16 dependency to BATMAN_ADV_BLA
From: Simon Wunderlich @ 2018-11-14 14:08 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The commit ced72933a5e8 ("batman-adv: use CRC32C instead of CRC16 in TT
code") switched the translation table code from crc16 to crc32c. The
(optional) bridge loop avoidance code is the only user of this function.
batman-adv should only select CRC16 when it is actually using it.
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/Kconfig | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/batman-adv/Kconfig b/net/batman-adv/Kconfig
index 082e96060bc2..d6b94559f888 100644
--- a/net/batman-adv/Kconfig
+++ b/net/batman-adv/Kconfig
@@ -22,7 +22,6 @@
config BATMAN_ADV
tristate "B.A.T.M.A.N. Advanced Meshing Protocol"
depends on NET
- select CRC16
select LIBCRC32C
help
B.A.T.M.A.N. (better approach to mobile ad-hoc networking) is
@@ -48,6 +47,7 @@ config BATMAN_ADV_BATMAN_V
config BATMAN_ADV_BLA
bool "Bridge Loop Avoidance"
depends on BATMAN_ADV && INET
+ select CRC16
default y
help
This option enables BLA (Bridge Loop Avoidance), a mechanism
--
2.11.0
^ permalink raw reply related
* [PATCH 14/16] batman-adv: Add inconsistent multicast netlink dump detection
From: Simon Wunderlich @ 2018-11-14 14:08 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The netlink dump functionality transfers a large number of entries from the
kernel to userspace. It is rather likely that the transfer has to
interrupted and later continued. During that time, it can happen that
either new entries are added or removed. The userspace could than either
receive some entries multiple times or miss entries.
Commit 670dc2833d14 ("netlink: advertise incomplete dumps") introduced a
mechanism to inform userspace about this problem. Userspace can then decide
whether it is necessary or not to retry dumping the information again.
The netlink dump functions have to be switched to exclusive locks to avoid
changes while the current message is prepared. The already existing
generation sequence counter from the hash helper can be used for this
simple hash.
Reported-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/multicast.c | 51 +++++++++++++++++++++++++---------------------
1 file changed, 28 insertions(+), 23 deletions(-)
diff --git a/net/batman-adv/multicast.c b/net/batman-adv/multicast.c
index 86725d792e15..69244e4598f5 100644
--- a/net/batman-adv/multicast.c
+++ b/net/batman-adv/multicast.c
@@ -1365,22 +1365,26 @@ int batadv_mcast_mesh_info_put(struct sk_buff *msg,
* to a netlink socket
* @msg: buffer for the message
* @portid: netlink port
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @orig_node: originator to dump the multicast flags of
*
* Return: 0 or error code.
*/
static int
-batadv_mcast_flags_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
+batadv_mcast_flags_dump_entry(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_orig_node *orig_node)
{
void *hdr;
- hdr = genlmsg_put(msg, portid, seq, &batadv_netlink_family,
- NLM_F_MULTI, BATADV_CMD_GET_MCAST_FLAGS);
+ hdr = genlmsg_put(msg, portid, cb->nlh->nlmsg_seq,
+ &batadv_netlink_family, NLM_F_MULTI,
+ BATADV_CMD_GET_MCAST_FLAGS);
if (!hdr)
return -ENOBUFS;
+ genl_dump_check_consistent(cb, hdr);
+
if (nla_put(msg, BATADV_ATTR_ORIG_ADDRESS, ETH_ALEN,
orig_node->orig)) {
genlmsg_cancel(msg, hdr);
@@ -1405,21 +1409,26 @@ batadv_mcast_flags_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
* table to a netlink socket
* @msg: buffer for the message
* @portid: netlink port
- * @seq: Sequence number of netlink message
- * @head: bucket to dump
+ * @cb: Control block containing additional options
+ * @hash: hash to dump
+ * @bucket: bucket index to dump
* @idx_skip: How many entries to skip
*
* Return: 0 or error code.
*/
static int
-batadv_mcast_flags_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
- struct hlist_head *head, long *idx_skip)
+batadv_mcast_flags_dump_bucket(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
+ struct batadv_hashtable *hash,
+ unsigned int bucket, long *idx_skip)
{
struct batadv_orig_node *orig_node;
long idx = 0;
- rcu_read_lock();
- hlist_for_each_entry_rcu(orig_node, head, hash_entry) {
+ spin_lock_bh(&hash->list_locks[bucket]);
+ cb->seq = atomic_read(&hash->generation) << 1 | 1;
+
+ hlist_for_each_entry(orig_node, &hash->table[bucket], hash_entry) {
if (!test_bit(BATADV_ORIG_CAPA_HAS_MCAST,
&orig_node->capa_initialized))
continue;
@@ -1427,9 +1436,8 @@ batadv_mcast_flags_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
if (idx < *idx_skip)
goto skip;
- if (batadv_mcast_flags_dump_entry(msg, portid, seq,
- orig_node)) {
- rcu_read_unlock();
+ if (batadv_mcast_flags_dump_entry(msg, portid, cb, orig_node)) {
+ spin_unlock_bh(&hash->list_locks[bucket]);
*idx_skip = idx;
return -EMSGSIZE;
@@ -1438,7 +1446,7 @@ batadv_mcast_flags_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
skip:
idx++;
}
- rcu_read_unlock();
+ spin_unlock_bh(&hash->list_locks[bucket]);
return 0;
}
@@ -1447,7 +1455,7 @@ batadv_mcast_flags_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
* __batadv_mcast_flags_dump() - dump multicast flags table to a netlink socket
* @msg: buffer for the message
* @portid: netlink port
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @bat_priv: the bat priv with all the soft interface information
* @bucket: current bucket to dump
* @idx: index in current bucket to the next entry to dump
@@ -1455,19 +1463,17 @@ batadv_mcast_flags_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
* Return: 0 or error code.
*/
static int
-__batadv_mcast_flags_dump(struct sk_buff *msg, u32 portid, u32 seq,
+__batadv_mcast_flags_dump(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_priv *bat_priv, long *bucket, long *idx)
{
struct batadv_hashtable *hash = bat_priv->orig_hash;
long bucket_tmp = *bucket;
- struct hlist_head *head;
long idx_tmp = *idx;
while (bucket_tmp < hash->size) {
- head = &hash->table[bucket_tmp];
-
- if (batadv_mcast_flags_dump_bucket(msg, portid, seq, head,
- &idx_tmp))
+ if (batadv_mcast_flags_dump_bucket(msg, portid, cb, hash,
+ *bucket, &idx_tmp))
break;
bucket_tmp++;
@@ -1550,8 +1556,7 @@ int batadv_mcast_flags_dump(struct sk_buff *msg, struct netlink_callback *cb)
return ret;
bat_priv = netdev_priv(primary_if->soft_iface);
- ret = __batadv_mcast_flags_dump(msg, portid, cb->nlh->nlmsg_seq,
- bat_priv, bucket, idx);
+ ret = __batadv_mcast_flags_dump(msg, portid, cb, bat_priv, bucket, idx);
batadv_hardif_put(primary_if);
return ret;
--
2.11.0
^ permalink raw reply related
* [PATCH 13/16] batman-adv: Add inconsistent local TT netlink dump detection
From: Simon Wunderlich @ 2018-11-14 14:08 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The netlink dump functionality transfers a large number of entries from the
kernel to userspace. It is rather likely that the transfer has to
interrupted and later continued. During that time, it can happen that
either new entries are added or removed. The userspace could than either
receive some entries multiple times or miss entries.
Commit 670dc2833d14 ("netlink: advertise incomplete dumps") introduced a
mechanism to inform userspace about this problem. Userspace can then decide
whether it is necessary or not to retry dumping the information again.
The netlink dump functions have to be switched to exclusive locks to avoid
changes while the current message is prepared. The already existing
generation sequence counter from the hash helper can be used for this
simple hash.
Reported-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/translation-table.c | 41 +++++++++++++++++++++-----------------
1 file changed, 23 insertions(+), 18 deletions(-)
diff --git a/net/batman-adv/translation-table.c b/net/batman-adv/translation-table.c
index d21624c44665..8dcd4968cde7 100644
--- a/net/batman-adv/translation-table.c
+++ b/net/batman-adv/translation-table.c
@@ -1145,14 +1145,15 @@ int batadv_tt_local_seq_print_text(struct seq_file *seq, void *offset)
* batadv_tt_local_dump_entry() - Dump one TT local entry into a message
* @msg :Netlink message to dump into
* @portid: Port making netlink request
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @bat_priv: The bat priv with all the soft interface information
* @common: tt local & tt global common data
*
* Return: Error code, or 0 on success
*/
static int
-batadv_tt_local_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
+batadv_tt_local_dump_entry(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_priv *bat_priv,
struct batadv_tt_common_entry *common)
{
@@ -1173,12 +1174,14 @@ batadv_tt_local_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
batadv_softif_vlan_put(vlan);
- hdr = genlmsg_put(msg, portid, seq, &batadv_netlink_family,
- NLM_F_MULTI,
+ hdr = genlmsg_put(msg, portid, cb->nlh->nlmsg_seq,
+ &batadv_netlink_family, NLM_F_MULTI,
BATADV_CMD_GET_TRANSTABLE_LOCAL);
if (!hdr)
return -ENOBUFS;
+ genl_dump_check_consistent(cb, hdr);
+
if (nla_put(msg, BATADV_ATTR_TT_ADDRESS, ETH_ALEN, common->addr) ||
nla_put_u32(msg, BATADV_ATTR_TT_CRC32, crc) ||
nla_put_u16(msg, BATADV_ATTR_TT_VID, common->vid) ||
@@ -1201,34 +1204,39 @@ batadv_tt_local_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
* batadv_tt_local_dump_bucket() - Dump one TT local bucket into a message
* @msg: Netlink message to dump into
* @portid: Port making netlink request
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @bat_priv: The bat priv with all the soft interface information
- * @head: Pointer to the list containing the local tt entries
+ * @hash: hash to dump
+ * @bucket: bucket index to dump
* @idx_s: Number of entries to skip
*
* Return: Error code, or 0 on success
*/
static int
-batadv_tt_local_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
+batadv_tt_local_dump_bucket(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_priv *bat_priv,
- struct hlist_head *head, int *idx_s)
+ struct batadv_hashtable *hash, unsigned int bucket,
+ int *idx_s)
{
struct batadv_tt_common_entry *common;
int idx = 0;
- rcu_read_lock();
- hlist_for_each_entry_rcu(common, head, hash_entry) {
+ spin_lock_bh(&hash->list_locks[bucket]);
+ cb->seq = atomic_read(&hash->generation) << 1 | 1;
+
+ hlist_for_each_entry(common, &hash->table[bucket], hash_entry) {
if (idx++ < *idx_s)
continue;
- if (batadv_tt_local_dump_entry(msg, portid, seq, bat_priv,
+ if (batadv_tt_local_dump_entry(msg, portid, cb, bat_priv,
common)) {
- rcu_read_unlock();
+ spin_unlock_bh(&hash->list_locks[bucket]);
*idx_s = idx - 1;
return -EMSGSIZE;
}
}
- rcu_read_unlock();
+ spin_unlock_bh(&hash->list_locks[bucket]);
*idx_s = 0;
return 0;
@@ -1248,7 +1256,6 @@ int batadv_tt_local_dump(struct sk_buff *msg, struct netlink_callback *cb)
struct batadv_priv *bat_priv;
struct batadv_hard_iface *primary_if = NULL;
struct batadv_hashtable *hash;
- struct hlist_head *head;
int ret;
int ifindex;
int bucket = cb->args[0];
@@ -1276,10 +1283,8 @@ int batadv_tt_local_dump(struct sk_buff *msg, struct netlink_callback *cb)
hash = bat_priv->tt.local_hash;
while (bucket < hash->size) {
- head = &hash->table[bucket];
-
- if (batadv_tt_local_dump_bucket(msg, portid, cb->nlh->nlmsg_seq,
- bat_priv, head, &idx))
+ if (batadv_tt_local_dump_bucket(msg, portid, cb, bat_priv,
+ hash, bucket, &idx))
break;
bucket++;
--
2.11.0
^ permalink raw reply related
* [PATCH 12/16] batman-adv: Add inconsistent dat netlink dump detection
From: Simon Wunderlich @ 2018-11-14 14:08 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The netlink dump functionality transfers a large number of entries from the
kernel to userspace. It is rather likely that the transfer has to
interrupted and later continued. During that time, it can happen that
either new entries are added or removed. The userspace could than either
receive some entries multiple times or miss entries.
Commit 670dc2833d14 ("netlink: advertise incomplete dumps") introduced a
mechanism to inform userspace about this problem. Userspace can then decide
whether it is necessary or not to retry dumping the information again.
The netlink dump functions have to be switched to exclusive locks to avoid
changes while the current message is prepared. The already existing
generation sequence counter from the hash helper can be used for this
simple hash.
Reported-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/distributed-arp-table.c | 42 +++++++++++++++++++---------------
1 file changed, 23 insertions(+), 19 deletions(-)
diff --git a/net/batman-adv/distributed-arp-table.c b/net/batman-adv/distributed-arp-table.c
index a60bacf7120b..b9ffe1826527 100644
--- a/net/batman-adv/distributed-arp-table.c
+++ b/net/batman-adv/distributed-arp-table.c
@@ -863,23 +863,27 @@ int batadv_dat_cache_seq_print_text(struct seq_file *seq, void *offset)
* netlink socket
* @msg: buffer for the message
* @portid: netlink port
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @dat_entry: entry to dump
*
* Return: 0 or error code.
*/
static int
-batadv_dat_cache_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
+batadv_dat_cache_dump_entry(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_dat_entry *dat_entry)
{
int msecs;
void *hdr;
- hdr = genlmsg_put(msg, portid, seq, &batadv_netlink_family,
- NLM_F_MULTI, BATADV_CMD_GET_DAT_CACHE);
+ hdr = genlmsg_put(msg, portid, cb->nlh->nlmsg_seq,
+ &batadv_netlink_family, NLM_F_MULTI,
+ BATADV_CMD_GET_DAT_CACHE);
if (!hdr)
return -ENOBUFS;
+ genl_dump_check_consistent(cb, hdr);
+
msecs = jiffies_to_msecs(jiffies - dat_entry->last_update);
if (nla_put_in_addr(msg, BATADV_ATTR_DAT_CACHE_IP4ADDRESS,
@@ -901,27 +905,31 @@ batadv_dat_cache_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
* a netlink socket
* @msg: buffer for the message
* @portid: netlink port
- * @seq: Sequence number of netlink message
- * @head: bucket to dump
+ * @cb: Control block containing additional options
+ * @hash: hash to dump
+ * @bucket: bucket index to dump
* @idx_skip: How many entries to skip
*
* Return: 0 or error code.
*/
static int
-batadv_dat_cache_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
- struct hlist_head *head, int *idx_skip)
+batadv_dat_cache_dump_bucket(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
+ struct batadv_hashtable *hash, unsigned int bucket,
+ int *idx_skip)
{
struct batadv_dat_entry *dat_entry;
int idx = 0;
- rcu_read_lock();
- hlist_for_each_entry_rcu(dat_entry, head, hash_entry) {
+ spin_lock_bh(&hash->list_locks[bucket]);
+ cb->seq = atomic_read(&hash->generation) << 1 | 1;
+
+ hlist_for_each_entry(dat_entry, &hash->table[bucket], hash_entry) {
if (idx < *idx_skip)
goto skip;
- if (batadv_dat_cache_dump_entry(msg, portid, seq,
- dat_entry)) {
- rcu_read_unlock();
+ if (batadv_dat_cache_dump_entry(msg, portid, cb, dat_entry)) {
+ spin_unlock_bh(&hash->list_locks[bucket]);
*idx_skip = idx;
return -EMSGSIZE;
@@ -930,7 +938,7 @@ batadv_dat_cache_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
skip:
idx++;
}
- rcu_read_unlock();
+ spin_unlock_bh(&hash->list_locks[bucket]);
return 0;
}
@@ -951,7 +959,6 @@ int batadv_dat_cache_dump(struct sk_buff *msg, struct netlink_callback *cb)
struct batadv_hashtable *hash;
struct batadv_priv *bat_priv;
int bucket = cb->args[0];
- struct hlist_head *head;
int idx = cb->args[1];
int ifindex;
int ret = 0;
@@ -977,10 +984,7 @@ int batadv_dat_cache_dump(struct sk_buff *msg, struct netlink_callback *cb)
}
while (bucket < hash->size) {
- head = &hash->table[bucket];
-
- if (batadv_dat_cache_dump_bucket(msg, portid,
- cb->nlh->nlmsg_seq, head,
+ if (batadv_dat_cache_dump_bucket(msg, portid, cb, hash, bucket,
&idx))
break;
--
2.11.0
^ permalink raw reply related
* [PATCH 11/16] batman-adv: Add inconsistent claim netlink dump detection
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The netlink dump functionality transfers a large number of entries from the
kernel to userspace. It is rather likely that the transfer has to
interrupted and later continued. During that time, it can happen that
either new entries are added or removed. The userspace could than either
receive some entries multiple times or miss entries.
Commit 670dc2833d14 ("netlink: advertise incomplete dumps") introduced a
mechanism to inform userspace about this problem. Userspace can then decide
whether it is necessary or not to retry dumping the information again.
The netlink dump functions have to be switched to exclusive locks to avoid
changes while the current message is prepared. The already existing
generation sequence counter from the hash helper can be used for this
simple hash.
Reported-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/bridge_loop_avoidance.c | 41 +++++++++++++++++++---------------
1 file changed, 23 insertions(+), 18 deletions(-)
diff --git a/net/batman-adv/bridge_loop_avoidance.c b/net/batman-adv/bridge_loop_avoidance.c
index 9f3747346d29..5fdde2947802 100644
--- a/net/batman-adv/bridge_loop_avoidance.c
+++ b/net/batman-adv/bridge_loop_avoidance.c
@@ -2094,14 +2094,15 @@ int batadv_bla_claim_table_seq_print_text(struct seq_file *seq, void *offset)
* to a netlink socket
* @msg: buffer for the message
* @portid: netlink port
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @primary_if: primary interface
* @claim: entry to dump
*
* Return: 0 or error code.
*/
static int
-batadv_bla_claim_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
+batadv_bla_claim_dump_entry(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_hard_iface *primary_if,
struct batadv_bla_claim *claim)
{
@@ -2111,13 +2112,16 @@ batadv_bla_claim_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
void *hdr;
int ret = -EINVAL;
- hdr = genlmsg_put(msg, portid, seq, &batadv_netlink_family,
- NLM_F_MULTI, BATADV_CMD_GET_BLA_CLAIM);
+ hdr = genlmsg_put(msg, portid, cb->nlh->nlmsg_seq,
+ &batadv_netlink_family, NLM_F_MULTI,
+ BATADV_CMD_GET_BLA_CLAIM);
if (!hdr) {
ret = -ENOBUFS;
goto out;
}
+ genl_dump_check_consistent(cb, hdr);
+
is_own = batadv_compare_eth(claim->backbone_gw->orig,
primary_addr);
@@ -2153,28 +2157,33 @@ batadv_bla_claim_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
* to a netlink socket
* @msg: buffer for the message
* @portid: netlink port
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @primary_if: primary interface
- * @head: bucket to dump
+ * @hash: hash to dump
+ * @bucket: bucket index to dump
* @idx_skip: How many entries to skip
*
* Return: always 0.
*/
static int
-batadv_bla_claim_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
+batadv_bla_claim_dump_bucket(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_hard_iface *primary_if,
- struct hlist_head *head, int *idx_skip)
+ struct batadv_hashtable *hash, unsigned int bucket,
+ int *idx_skip)
{
struct batadv_bla_claim *claim;
int idx = 0;
int ret = 0;
- rcu_read_lock();
- hlist_for_each_entry_rcu(claim, head, hash_entry) {
+ spin_lock_bh(&hash->list_locks[bucket]);
+ cb->seq = atomic_read(&hash->generation) << 1 | 1;
+
+ hlist_for_each_entry(claim, &hash->table[bucket], hash_entry) {
if (idx++ < *idx_skip)
continue;
- ret = batadv_bla_claim_dump_entry(msg, portid, seq,
+ ret = batadv_bla_claim_dump_entry(msg, portid, cb,
primary_if, claim);
if (ret) {
*idx_skip = idx - 1;
@@ -2184,7 +2193,7 @@ batadv_bla_claim_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
*idx_skip = 0;
unlock:
- rcu_read_unlock();
+ spin_unlock_bh(&hash->list_locks[bucket]);
return ret;
}
@@ -2204,7 +2213,6 @@ int batadv_bla_claim_dump(struct sk_buff *msg, struct netlink_callback *cb)
struct batadv_hashtable *hash;
struct batadv_priv *bat_priv;
int bucket = cb->args[0];
- struct hlist_head *head;
int idx = cb->args[1];
int ifindex;
int ret = 0;
@@ -2230,11 +2238,8 @@ int batadv_bla_claim_dump(struct sk_buff *msg, struct netlink_callback *cb)
}
while (bucket < hash->size) {
- head = &hash->table[bucket];
-
- if (batadv_bla_claim_dump_bucket(msg, portid,
- cb->nlh->nlmsg_seq,
- primary_if, head, &idx))
+ if (batadv_bla_claim_dump_bucket(msg, portid, cb, primary_if,
+ hash, bucket, &idx))
break;
bucket++;
}
--
2.11.0
^ permalink raw reply related
* [PATCH 10/16] batman-adv: Add inconsistent backbone netlink dump detection
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The netlink dump functionality transfers a large number of entries from the
kernel to userspace. It is rather likely that the transfer has to
interrupted and later continued. During that time, it can happen that
either new entries are added or removed. The userspace could than either
receive some entries multiple times or miss entries.
Commit 670dc2833d14 ("netlink: advertise incomplete dumps") introduced a
mechanism to inform userspace about this problem. Userspace can then decide
whether it is necessary or not to retry dumping the information again.
The netlink dump functions have to be switched to exclusive locks to avoid
changes while the current message is prepared. The already existing
generation sequence counter from the hash helper can be used for this
simple hash.
Reported-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/bridge_loop_avoidance.c | 41 +++++++++++++++++++---------------
1 file changed, 23 insertions(+), 18 deletions(-)
diff --git a/net/batman-adv/bridge_loop_avoidance.c b/net/batman-adv/bridge_loop_avoidance.c
index 5f1aeeded0e3..9f3747346d29 100644
--- a/net/batman-adv/bridge_loop_avoidance.c
+++ b/net/batman-adv/bridge_loop_avoidance.c
@@ -2325,14 +2325,15 @@ int batadv_bla_backbone_table_seq_print_text(struct seq_file *seq, void *offset)
* netlink socket
* @msg: buffer for the message
* @portid: netlink port
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @primary_if: primary interface
* @backbone_gw: entry to dump
*
* Return: 0 or error code.
*/
static int
-batadv_bla_backbone_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
+batadv_bla_backbone_dump_entry(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_hard_iface *primary_if,
struct batadv_bla_backbone_gw *backbone_gw)
{
@@ -2343,13 +2344,16 @@ batadv_bla_backbone_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
void *hdr;
int ret = -EINVAL;
- hdr = genlmsg_put(msg, portid, seq, &batadv_netlink_family,
- NLM_F_MULTI, BATADV_CMD_GET_BLA_BACKBONE);
+ hdr = genlmsg_put(msg, portid, cb->nlh->nlmsg_seq,
+ &batadv_netlink_family, NLM_F_MULTI,
+ BATADV_CMD_GET_BLA_BACKBONE);
if (!hdr) {
ret = -ENOBUFS;
goto out;
}
+ genl_dump_check_consistent(cb, hdr);
+
is_own = batadv_compare_eth(backbone_gw->orig, primary_addr);
spin_lock_bh(&backbone_gw->crc_lock);
@@ -2386,28 +2390,33 @@ batadv_bla_backbone_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
* a netlink socket
* @msg: buffer for the message
* @portid: netlink port
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @primary_if: primary interface
- * @head: bucket to dump
+ * @hash: hash to dump
+ * @bucket: bucket index to dump
* @idx_skip: How many entries to skip
*
* Return: always 0.
*/
static int
-batadv_bla_backbone_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
+batadv_bla_backbone_dump_bucket(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_hard_iface *primary_if,
- struct hlist_head *head, int *idx_skip)
+ struct batadv_hashtable *hash,
+ unsigned int bucket, int *idx_skip)
{
struct batadv_bla_backbone_gw *backbone_gw;
int idx = 0;
int ret = 0;
- rcu_read_lock();
- hlist_for_each_entry_rcu(backbone_gw, head, hash_entry) {
+ spin_lock_bh(&hash->list_locks[bucket]);
+ cb->seq = atomic_read(&hash->generation) << 1 | 1;
+
+ hlist_for_each_entry(backbone_gw, &hash->table[bucket], hash_entry) {
if (idx++ < *idx_skip)
continue;
- ret = batadv_bla_backbone_dump_entry(msg, portid, seq,
+ ret = batadv_bla_backbone_dump_entry(msg, portid, cb,
primary_if, backbone_gw);
if (ret) {
*idx_skip = idx - 1;
@@ -2417,7 +2426,7 @@ batadv_bla_backbone_dump_bucket(struct sk_buff *msg, u32 portid, u32 seq,
*idx_skip = 0;
unlock:
- rcu_read_unlock();
+ spin_unlock_bh(&hash->list_locks[bucket]);
return ret;
}
@@ -2437,7 +2446,6 @@ int batadv_bla_backbone_dump(struct sk_buff *msg, struct netlink_callback *cb)
struct batadv_hashtable *hash;
struct batadv_priv *bat_priv;
int bucket = cb->args[0];
- struct hlist_head *head;
int idx = cb->args[1];
int ifindex;
int ret = 0;
@@ -2463,11 +2471,8 @@ int batadv_bla_backbone_dump(struct sk_buff *msg, struct netlink_callback *cb)
}
while (bucket < hash->size) {
- head = &hash->table[bucket];
-
- if (batadv_bla_backbone_dump_bucket(msg, portid,
- cb->nlh->nlmsg_seq,
- primary_if, head, &idx))
+ if (batadv_bla_backbone_dump_bucket(msg, portid, cb, primary_if,
+ hash, bucket, &idx))
break;
bucket++;
}
--
2.11.0
^ permalink raw reply related
* [PATCH 09/16] batman-adv: Store modification counter via hash helpers
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
Multiple datastructures use the hash helper functions to add and remove
entries from the simple hlist based hashes. These are often also dumped to
userspace via netlink and thus should have a generation sequence counter.
Reported-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/hash.c | 2 ++
net/batman-adv/hash.h | 6 ++++++
2 files changed, 8 insertions(+)
diff --git a/net/batman-adv/hash.c b/net/batman-adv/hash.c
index 7b49e4001778..9194f4d891b1 100644
--- a/net/batman-adv/hash.c
+++ b/net/batman-adv/hash.c
@@ -32,6 +32,8 @@ static void batadv_hash_init(struct batadv_hashtable *hash)
INIT_HLIST_HEAD(&hash->table[i]);
spin_lock_init(&hash->list_locks[i]);
}
+
+ atomic_set(&hash->generation, 0);
}
/**
diff --git a/net/batman-adv/hash.h b/net/batman-adv/hash.h
index 9490a7ca2ba6..0e36fa1c7c3e 100644
--- a/net/batman-adv/hash.h
+++ b/net/batman-adv/hash.h
@@ -21,6 +21,7 @@
#include "main.h"
+#include <linux/atomic.h>
#include <linux/compiler.h>
#include <linux/list.h>
#include <linux/rculist.h>
@@ -58,6 +59,9 @@ struct batadv_hashtable {
/** @size: size of hashtable */
u32 size;
+
+ /** @generation: current (generation) sequence number */
+ atomic_t generation;
};
/* allocates and clears the hash */
@@ -112,6 +116,7 @@ static inline int batadv_hash_add(struct batadv_hashtable *hash,
/* no duplicate found in list, add new element */
hlist_add_head_rcu(data_node, head);
+ atomic_inc(&hash->generation);
ret = 0;
@@ -154,6 +159,7 @@ static inline void *batadv_hash_remove(struct batadv_hashtable *hash,
data_save = node;
hlist_del_rcu(node);
+ atomic_inc(&hash->generation);
break;
}
spin_unlock_bh(&hash->list_locks[index]);
--
2.11.0
^ permalink raw reply related
* [PATCH 08/16] batman-adv: Add inconsistent hardif netlink dump detection
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The netlink dump functionality transfers a large number of entries from the
kernel to userspace. It is rather likely that the transfer has to
interrupted and later continued. During that time, it can happen that
either new entries are added or removed. The userspace could than either
receive some entries multiple times or miss entries.
Commit 670dc2833d14 ("netlink: advertise incomplete dumps") introduced a
mechanism to inform userspace about this problem. Userspace can then decide
whether it is necessary or not to retry dumping the information again.
The netlink dump functions have to be switched to exclusive locks to avoid
changes while the current message is prepared. And an external generation
sequence counter is introduced which tracks all modifications of the list.
Reported-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/hard-interface.c | 3 +++
net/batman-adv/main.c | 1 +
net/batman-adv/main.h | 1 +
net/batman-adv/netlink.c | 24 ++++++++++++++----------
4 files changed, 19 insertions(+), 10 deletions(-)
diff --git a/net/batman-adv/hard-interface.c b/net/batman-adv/hard-interface.c
index 781c5b6e6e8e..508f4416dfc9 100644
--- a/net/batman-adv/hard-interface.c
+++ b/net/batman-adv/hard-interface.c
@@ -951,6 +951,7 @@ batadv_hardif_add_interface(struct net_device *net_dev)
batadv_check_known_mac_addr(hard_iface->net_dev);
kref_get(&hard_iface->refcount);
list_add_tail_rcu(&hard_iface->list, &batadv_hardif_list);
+ batadv_hardif_generation++;
return hard_iface;
@@ -993,6 +994,7 @@ void batadv_hardif_remove_interfaces(void)
list_for_each_entry_safe(hard_iface, hard_iface_tmp,
&batadv_hardif_list, list) {
list_del_rcu(&hard_iface->list);
+ batadv_hardif_generation++;
batadv_hardif_remove_interface(hard_iface);
}
rtnl_unlock();
@@ -1054,6 +1056,7 @@ static int batadv_hard_if_event(struct notifier_block *this,
case NETDEV_UNREGISTER:
case NETDEV_PRE_TYPE_CHANGE:
list_del_rcu(&hard_iface->list);
+ batadv_hardif_generation++;
batadv_hardif_remove_interface(hard_iface);
break;
diff --git a/net/batman-adv/main.c b/net/batman-adv/main.c
index c75e47826949..d1ed839fd32b 100644
--- a/net/batman-adv/main.c
+++ b/net/batman-adv/main.c
@@ -74,6 +74,7 @@
* list traversals just rcu-locked
*/
struct list_head batadv_hardif_list;
+unsigned int batadv_hardif_generation;
static int (*batadv_rx_handler[256])(struct sk_buff *skb,
struct batadv_hard_iface *recv_if);
diff --git a/net/batman-adv/main.h b/net/batman-adv/main.h
index b68a41190eb0..b572066325e4 100644
--- a/net/batman-adv/main.h
+++ b/net/batman-adv/main.h
@@ -247,6 +247,7 @@ static inline int batadv_print_vid(unsigned short vid)
}
extern struct list_head batadv_hardif_list;
+extern unsigned int batadv_hardif_generation;
extern unsigned char batadv_broadcast_addr[];
extern struct workqueue_struct *batadv_event_workqueue;
diff --git a/net/batman-adv/netlink.c b/net/batman-adv/netlink.c
index 0d9459b69bdb..2dc3304cee54 100644
--- a/net/batman-adv/netlink.c
+++ b/net/batman-adv/netlink.c
@@ -29,11 +29,11 @@
#include <linux/if_ether.h>
#include <linux/init.h>
#include <linux/kernel.h>
+#include <linux/list.h>
#include <linux/netdevice.h>
#include <linux/netlink.h>
#include <linux/printk.h>
-#include <linux/rculist.h>
-#include <linux/rcupdate.h>
+#include <linux/rtnetlink.h>
#include <linux/skbuff.h>
#include <linux/stddef.h>
#include <linux/types.h>
@@ -445,23 +445,27 @@ batadv_netlink_tp_meter_cancel(struct sk_buff *skb, struct genl_info *info)
* batadv_netlink_dump_hardif_entry() - Dump one hard interface into a message
* @msg: Netlink message to dump into
* @portid: Port making netlink request
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @hard_iface: Hard interface to dump
*
* Return: error code, or 0 on success
*/
static int
-batadv_netlink_dump_hardif_entry(struct sk_buff *msg, u32 portid, u32 seq,
+batadv_netlink_dump_hardif_entry(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_hard_iface *hard_iface)
{
struct net_device *net_dev = hard_iface->net_dev;
void *hdr;
- hdr = genlmsg_put(msg, portid, seq, &batadv_netlink_family, NLM_F_MULTI,
+ hdr = genlmsg_put(msg, portid, cb->nlh->nlmsg_seq,
+ &batadv_netlink_family, NLM_F_MULTI,
BATADV_CMD_GET_HARDIFS);
if (!hdr)
return -EMSGSIZE;
+ genl_dump_check_consistent(cb, hdr);
+
if (nla_put_u32(msg, BATADV_ATTR_HARD_IFINDEX,
net_dev->ifindex) ||
nla_put_string(msg, BATADV_ATTR_HARD_IFNAME,
@@ -498,7 +502,6 @@ batadv_netlink_dump_hardifs(struct sk_buff *msg, struct netlink_callback *cb)
struct batadv_hard_iface *hard_iface;
int ifindex;
int portid = NETLINK_CB(cb->skb).portid;
- int seq = cb->nlh->nlmsg_seq;
int skip = cb->args[0];
int i = 0;
@@ -516,23 +519,24 @@ batadv_netlink_dump_hardifs(struct sk_buff *msg, struct netlink_callback *cb)
return -ENODEV;
}
- rcu_read_lock();
+ rtnl_lock();
+ cb->seq = batadv_hardif_generation << 1 | 1;
- list_for_each_entry_rcu(hard_iface, &batadv_hardif_list, list) {
+ list_for_each_entry(hard_iface, &batadv_hardif_list, list) {
if (hard_iface->soft_iface != soft_iface)
continue;
if (i++ < skip)
continue;
- if (batadv_netlink_dump_hardif_entry(msg, portid, seq,
+ if (batadv_netlink_dump_hardif_entry(msg, portid, cb,
hard_iface)) {
i--;
break;
}
}
- rcu_read_unlock();
+ rtnl_unlock();
dev_put(soft_iface);
--
2.11.0
^ permalink raw reply related
* [PATCH 07/16] batman-adv: Add inconsistent gateway netlink dump detection
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The netlink dump functionality transfers a large number of entries from the
kernel to userspace. It is rather likely that the transfer has to
interrupted and later continued. During that time, it can happen that
either new entries are added or removed. The userspace could than either
receive some entries multiple times or miss entries.
Commit 670dc2833d14 ("netlink: advertise incomplete dumps") introduced a
mechanism to inform userspace about this problem. Userspace can then decide
whether it is necessary or not to retry dumping the information again.
The netlink dump functions have to be switched to exclusive locks to avoid
changes while the current message is prepared. And an external generation
sequence counter is introduced which tracks all modifications of the list.
Reported-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/bat_iv_ogm.c | 24 +++++++++++++++---------
net/batman-adv/bat_v.c | 26 +++++++++++++++++---------
net/batman-adv/gateway_client.c | 3 +++
net/batman-adv/main.c | 2 ++
net/batman-adv/types.h | 5 ++++-
5 files changed, 41 insertions(+), 19 deletions(-)
diff --git a/net/batman-adv/bat_iv_ogm.c b/net/batman-adv/bat_iv_ogm.c
index 1d31ac84dec7..f97e566f0402 100644
--- a/net/batman-adv/bat_iv_ogm.c
+++ b/net/batman-adv/bat_iv_ogm.c
@@ -2584,13 +2584,14 @@ static void batadv_iv_gw_print(struct batadv_priv *bat_priv,
* batadv_iv_gw_dump_entry() - Dump a gateway into a message
* @msg: Netlink message to dump into
* @portid: Port making netlink request
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @bat_priv: The bat priv with all the soft interface information
* @gw_node: Gateway to be dumped
*
* Return: Error code, or 0 on success
*/
-static int batadv_iv_gw_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
+static int batadv_iv_gw_dump_entry(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_priv *bat_priv,
struct batadv_gw_node *gw_node)
{
@@ -2610,13 +2611,16 @@ static int batadv_iv_gw_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
curr_gw = batadv_gw_get_selected_gw_node(bat_priv);
- hdr = genlmsg_put(msg, portid, seq, &batadv_netlink_family,
- NLM_F_MULTI, BATADV_CMD_GET_GATEWAYS);
+ hdr = genlmsg_put(msg, portid, cb->nlh->nlmsg_seq,
+ &batadv_netlink_family, NLM_F_MULTI,
+ BATADV_CMD_GET_GATEWAYS);
if (!hdr) {
ret = -ENOBUFS;
goto out;
}
+ genl_dump_check_consistent(cb, hdr);
+
ret = -EMSGSIZE;
if (curr_gw == gw_node)
@@ -2667,13 +2671,15 @@ static void batadv_iv_gw_dump(struct sk_buff *msg, struct netlink_callback *cb,
int idx_skip = cb->args[0];
int idx = 0;
- rcu_read_lock();
- hlist_for_each_entry_rcu(gw_node, &bat_priv->gw.gateway_list, list) {
+ spin_lock_bh(&bat_priv->gw.list_lock);
+ cb->seq = bat_priv->gw.generation << 1 | 1;
+
+ hlist_for_each_entry(gw_node, &bat_priv->gw.gateway_list, list) {
if (idx++ < idx_skip)
continue;
- if (batadv_iv_gw_dump_entry(msg, portid, cb->nlh->nlmsg_seq,
- bat_priv, gw_node)) {
+ if (batadv_iv_gw_dump_entry(msg, portid, cb, bat_priv,
+ gw_node)) {
idx_skip = idx - 1;
goto unlock;
}
@@ -2681,7 +2687,7 @@ static void batadv_iv_gw_dump(struct sk_buff *msg, struct netlink_callback *cb,
idx_skip = idx;
unlock:
- rcu_read_unlock();
+ spin_unlock_bh(&bat_priv->gw.list_lock);
cb->args[0] = idx_skip;
}
diff --git a/net/batman-adv/bat_v.c b/net/batman-adv/bat_v.c
index 6baec4e68898..90e33f84d37a 100644
--- a/net/batman-adv/bat_v.c
+++ b/net/batman-adv/bat_v.c
@@ -27,11 +27,13 @@
#include <linux/jiffies.h>
#include <linux/kernel.h>
#include <linux/kref.h>
+#include <linux/list.h>
#include <linux/netdevice.h>
#include <linux/netlink.h>
#include <linux/rculist.h>
#include <linux/rcupdate.h>
#include <linux/seq_file.h>
+#include <linux/spinlock.h>
#include <linux/stddef.h>
#include <linux/types.h>
#include <linux/workqueue.h>
@@ -915,13 +917,14 @@ static void batadv_v_gw_print(struct batadv_priv *bat_priv,
* batadv_v_gw_dump_entry() - Dump a gateway into a message
* @msg: Netlink message to dump into
* @portid: Port making netlink request
- * @seq: Sequence number of netlink message
+ * @cb: Control block containing additional options
* @bat_priv: The bat priv with all the soft interface information
* @gw_node: Gateway to be dumped
*
* Return: Error code, or 0 on success
*/
-static int batadv_v_gw_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
+static int batadv_v_gw_dump_entry(struct sk_buff *msg, u32 portid,
+ struct netlink_callback *cb,
struct batadv_priv *bat_priv,
struct batadv_gw_node *gw_node)
{
@@ -941,13 +944,16 @@ static int batadv_v_gw_dump_entry(struct sk_buff *msg, u32 portid, u32 seq,
curr_gw = batadv_gw_get_selected_gw_node(bat_priv);
- hdr = genlmsg_put(msg, portid, seq, &batadv_netlink_family,
- NLM_F_MULTI, BATADV_CMD_GET_GATEWAYS);
+ hdr = genlmsg_put(msg, portid, cb->nlh->nlmsg_seq,
+ &batadv_netlink_family, NLM_F_MULTI,
+ BATADV_CMD_GET_GATEWAYS);
if (!hdr) {
ret = -ENOBUFS;
goto out;
}
+ genl_dump_check_consistent(cb, hdr);
+
ret = -EMSGSIZE;
if (curr_gw == gw_node) {
@@ -1018,13 +1024,15 @@ static void batadv_v_gw_dump(struct sk_buff *msg, struct netlink_callback *cb,
int idx_skip = cb->args[0];
int idx = 0;
- rcu_read_lock();
- hlist_for_each_entry_rcu(gw_node, &bat_priv->gw.gateway_list, list) {
+ spin_lock_bh(&bat_priv->gw.list_lock);
+ cb->seq = bat_priv->gw.generation << 1 | 1;
+
+ hlist_for_each_entry(gw_node, &bat_priv->gw.gateway_list, list) {
if (idx++ < idx_skip)
continue;
- if (batadv_v_gw_dump_entry(msg, portid, cb->nlh->nlmsg_seq,
- bat_priv, gw_node)) {
+ if (batadv_v_gw_dump_entry(msg, portid, cb, bat_priv,
+ gw_node)) {
idx_skip = idx - 1;
goto unlock;
}
@@ -1032,7 +1040,7 @@ static void batadv_v_gw_dump(struct sk_buff *msg, struct netlink_callback *cb,
idx_skip = idx;
unlock:
- rcu_read_unlock();
+ spin_unlock_bh(&bat_priv->gw.list_lock);
cb->args[0] = idx_skip;
}
diff --git a/net/batman-adv/gateway_client.c b/net/batman-adv/gateway_client.c
index 140c61a3f1ec..9d8e5eda2314 100644
--- a/net/batman-adv/gateway_client.c
+++ b/net/batman-adv/gateway_client.c
@@ -377,6 +377,7 @@ static void batadv_gw_node_add(struct batadv_priv *bat_priv,
kref_get(&gw_node->refcount);
hlist_add_head_rcu(&gw_node->list, &bat_priv->gw.gateway_list);
+ bat_priv->gw.generation++;
batadv_dbg(BATADV_DBG_BATMAN, bat_priv,
"Found new gateway %pM -> gw bandwidth: %u.%u/%u.%u MBit\n",
@@ -472,6 +473,7 @@ void batadv_gw_node_update(struct batadv_priv *bat_priv,
if (!hlist_unhashed(&gw_node->list)) {
hlist_del_init_rcu(&gw_node->list);
batadv_gw_node_put(gw_node);
+ bat_priv->gw.generation++;
}
spin_unlock_bh(&bat_priv->gw.list_lock);
@@ -518,6 +520,7 @@ void batadv_gw_node_free(struct batadv_priv *bat_priv)
&bat_priv->gw.gateway_list, list) {
hlist_del_init_rcu(&gw_node->list);
batadv_gw_node_put(gw_node);
+ bat_priv->gw.generation++;
}
spin_unlock_bh(&bat_priv->gw.list_lock);
}
diff --git a/net/batman-adv/main.c b/net/batman-adv/main.c
index 69c0d85bceb3..c75e47826949 100644
--- a/net/batman-adv/main.c
+++ b/net/batman-adv/main.c
@@ -186,6 +186,8 @@ int batadv_mesh_init(struct net_device *soft_iface)
INIT_HLIST_HEAD(&bat_priv->softif_vlan_list);
INIT_HLIST_HEAD(&bat_priv->tp_list);
+ bat_priv->gw.generation = 0;
+
ret = batadv_v_mesh_init(bat_priv);
if (ret < 0)
goto err;
diff --git a/net/batman-adv/types.h b/net/batman-adv/types.h
index 45b5592de816..cbe17da36fcb 100644
--- a/net/batman-adv/types.h
+++ b/net/batman-adv/types.h
@@ -1096,12 +1096,15 @@ struct batadv_priv_gw {
/** @gateway_list: list of available gateway nodes */
struct hlist_head gateway_list;
- /** @list_lock: lock protecting gateway_list & curr_gw */
+ /** @list_lock: lock protecting gateway_list, curr_gw, generation */
spinlock_t list_lock;
/** @curr_gw: pointer to currently selected gateway node */
struct batadv_gw_node __rcu *curr_gw;
+ /** @generation: current (generation) sequence number */
+ unsigned int generation;
+
/**
* @mode: gateway operation: off, client or server (see batadv_gw_modes)
*/
--
2.11.0
^ permalink raw reply related
* [PATCH 05/16] batman-adv: Allow to use BATMAN_ADV_DEBUG without BATMAN_ADV_DEBUGFS
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The BATMAN_ADV_DEBUGFS portion of batman-adv is marked as deprecated. Thus
all required functionality should be available without it. The debug log
was already modified to also output via the kernel tracing function but
still retained its BATMAN_ADV_DEBUGFS functionality.
Separate the entry point for the debug log from the debugfs portions to
make it possible to build with BATMAN_ADV_DEBUG and without
BATMAN_ADV_DEBUGFS.
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/Kconfig | 2 +-
net/batman-adv/log.c | 60 +++++++++++++++++++++++++++-----------------------
2 files changed, 34 insertions(+), 28 deletions(-)
diff --git a/net/batman-adv/Kconfig b/net/batman-adv/Kconfig
index f75816f58107..7b0e1fcddfa7 100644
--- a/net/batman-adv/Kconfig
+++ b/net/batman-adv/Kconfig
@@ -100,7 +100,7 @@ config BATMAN_ADV_DEBUGFS
config BATMAN_ADV_DEBUG
bool "B.A.T.M.A.N. debugging"
- depends on BATMAN_ADV_DEBUGFS
+ depends on BATMAN_ADV
help
This is an option for use by developers; most people should
say N here. This enables compilation of support for
diff --git a/net/batman-adv/log.c b/net/batman-adv/log.c
index 6beb5f067810..02e55b78132f 100644
--- a/net/batman-adv/log.c
+++ b/net/batman-adv/log.c
@@ -43,6 +43,8 @@
#include "debugfs.h"
#include "trace.h"
+#ifdef CONFIG_BATMAN_ADV_DEBUGFS
+
#define BATADV_LOG_BUFF_MASK (batadv_log_buff_len - 1)
static const int batadv_log_buff_len = BATADV_LOG_BUF_LEN;
@@ -92,33 +94,6 @@ static int batadv_fdebug_log(struct batadv_priv_debug_log *debug_log,
return 0;
}
-/**
- * batadv_debug_log() - Add debug log entry
- * @bat_priv: the bat priv with all the soft interface information
- * @fmt: format string
- *
- * Return: 0 on success or negative error number in case of failure
- */
-int batadv_debug_log(struct batadv_priv *bat_priv, const char *fmt, ...)
-{
- struct va_format vaf;
- va_list args;
-
- va_start(args, fmt);
-
- vaf.fmt = fmt;
- vaf.va = &args;
-
- batadv_fdebug_log(bat_priv->debug_log, "[%10u] %pV",
- jiffies_to_msecs(jiffies), &vaf);
-
- trace_batadv_dbg(bat_priv, &vaf);
-
- va_end(args);
-
- return 0;
-}
-
static int batadv_log_open(struct inode *inode, struct file *file)
{
if (!try_module_get(THIS_MODULE))
@@ -259,3 +234,34 @@ void batadv_debug_log_cleanup(struct batadv_priv *bat_priv)
kfree(bat_priv->debug_log);
bat_priv->debug_log = NULL;
}
+
+#endif /* CONFIG_BATMAN_ADV_DEBUGFS */
+
+/**
+ * batadv_debug_log() - Add debug log entry
+ * @bat_priv: the bat priv with all the soft interface information
+ * @fmt: format string
+ *
+ * Return: 0 on success or negative error number in case of failure
+ */
+int batadv_debug_log(struct batadv_priv *bat_priv, const char *fmt, ...)
+{
+ struct va_format vaf;
+ va_list args;
+
+ va_start(args, fmt);
+
+ vaf.fmt = fmt;
+ vaf.va = &args;
+
+#ifdef CONFIG_BATMAN_ADV_DEBUGFS
+ batadv_fdebug_log(bat_priv->debug_log, "[%10u] %pV",
+ jiffies_to_msecs(jiffies), &vaf);
+#endif
+
+ trace_batadv_dbg(bat_priv, &vaf);
+
+ va_end(args);
+
+ return 0;
+}
--
2.11.0
^ permalink raw reply related
* [PATCH 06/16] batman-adv: Fix description for BATMAN_ADV_DEBUG
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The debug messages of batman-adv are not printed to the kernel log at all
but can be stored (depending on the compile setting) in the tracing buffer
or the batadv specific log buffer. There is also no debug module parameter
but a batadv netdev specific log_level setting to enable/disable different
classes of debug messages at runtime.
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/Kconfig | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/net/batman-adv/Kconfig b/net/batman-adv/Kconfig
index 7b0e1fcddfa7..082e96060bc2 100644
--- a/net/batman-adv/Kconfig
+++ b/net/batman-adv/Kconfig
@@ -104,8 +104,9 @@ config BATMAN_ADV_DEBUG
help
This is an option for use by developers; most people should
say N here. This enables compilation of support for
- outputting debugging information to the kernel log. The
- output is controlled via the module parameter debug.
+ outputting debugging information to the debugfs log or tracing
+ buffer. The output is controlled via the batadv netdev specific
+ log_level setting.
config BATMAN_ADV_TRACING
bool "B.A.T.M.A.N. tracing support"
--
2.11.0
^ permalink raw reply related
* [PATCH 04/16] batman-adv: Improve includes for trace functionality
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The batadv_dbg trace event uses different functionality and datastructures
which are not directly associated with the trace infrastructure. It should
not be expected that the trace headers indirectly provide them and instead
include the required headers directly.
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/trace.c | 2 --
net/batman-adv/trace.h | 6 ++++++
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/net/batman-adv/trace.c b/net/batman-adv/trace.c
index 3d57f9981f25..8e1024217cff 100644
--- a/net/batman-adv/trace.c
+++ b/net/batman-adv/trace.c
@@ -16,7 +16,5 @@
* along with this program; if not, see <http://www.gnu.org/licenses/>.
*/
-#include <linux/module.h>
-
#define CREATE_TRACE_POINTS
#include "trace.h"
diff --git a/net/batman-adv/trace.h b/net/batman-adv/trace.h
index 3acda26a30ca..104784be94d7 100644
--- a/net/batman-adv/trace.h
+++ b/net/batman-adv/trace.h
@@ -21,7 +21,13 @@
#include "main.h"
+#include <linux/bug.h>
+#include <linux/kernel.h>
+#include <linux/netdevice.h>
+#include <linux/percpu.h>
+#include <linux/printk.h>
#include <linux/tracepoint.h>
+#include <linux/types.h>
#undef TRACE_SYSTEM
#define TRACE_SYSTEM batadv
--
2.11.0
^ permalink raw reply related
* [PATCH 03/16] batman-adv: Add includes for deprecation warning
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The commit 00caf6a2b318 ("batman-adv: Mark debugfs functionality as
deprecated") introduced various messages to inform the user about the
deprecation of the debugfs based functionality. The messages also include
the context/task in which this problem was observed.
The datastructures and functions to access this information require special
headers. These should be included directly instead of depending on a more
complex and fragile include chain.
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/debugfs.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/net/batman-adv/debugfs.c b/net/batman-adv/debugfs.c
index 8b608a2e2653..d4a7702e48d8 100644
--- a/net/batman-adv/debugfs.c
+++ b/net/batman-adv/debugfs.c
@@ -19,6 +19,7 @@
#include "debugfs.h"
#include "main.h"
+#include <asm/current.h>
#include <linux/dcache.h>
#include <linux/debugfs.h>
#include <linux/err.h>
@@ -27,6 +28,7 @@
#include <linux/fs.h>
#include <linux/netdevice.h>
#include <linux/printk.h>
+#include <linux/sched.h>
#include <linux/seq_file.h>
#include <linux/stat.h>
#include <linux/stddef.h>
--
2.11.0
^ permalink raw reply related
* [PATCH 02/16] batman-adv: Drop unused lockdep include
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Sven Eckelmann, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
From: Sven Eckelmann <sven@narfation.org>
The commit dee222c7b20c ("batman-adv: Move OGM rebroadcast stats to
orig_ifinfo") removed all used functionality of the include linux/lockdep.h
from batadv_iv_ogm.c.
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/bat_iv_ogm.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/net/batman-adv/bat_iv_ogm.c b/net/batman-adv/bat_iv_ogm.c
index d2227091029f..1d31ac84dec7 100644
--- a/net/batman-adv/bat_iv_ogm.c
+++ b/net/batman-adv/bat_iv_ogm.c
@@ -34,7 +34,6 @@
#include <linux/kernel.h>
#include <linux/kref.h>
#include <linux/list.h>
-#include <linux/lockdep.h>
#include <linux/netdevice.h>
#include <linux/netlink.h>
#include <linux/pkt_sched.h>
--
2.11.0
^ permalink raw reply related
* [PATCH 01/16] batman-adv: Start new development cycle
From: Simon Wunderlich @ 2018-11-14 14:07 UTC (permalink / raw)
To: davem; +Cc: netdev, b.a.t.m.a.n, Simon Wunderlich
In-Reply-To: <20181114140804.18381-1-sw@simonwunderlich.de>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
---
net/batman-adv/main.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/batman-adv/main.h b/net/batman-adv/main.h
index 2002b70e18db..b68a41190eb0 100644
--- a/net/batman-adv/main.h
+++ b/net/batman-adv/main.h
@@ -25,7 +25,7 @@
#define BATADV_DRIVER_DEVICE "batman-adv"
#ifndef BATADV_SOURCE_VERSION
-#define BATADV_SOURCE_VERSION "2018.4"
+#define BATADV_SOURCE_VERSION "2019.0"
#endif
/* B.A.T.M.A.N. parameters */
--
2.11.0
^ permalink raw reply related
* [PATCH] rhashtable: detect when object movement between tables might have invalidated a lookup
From: NeilBrown @ 2018-11-14 23:32 UTC (permalink / raw)
To: David Miller, herbert, tgraf; +Cc: netdev, linux-kernel, eric.dumazet
[-- Attachment #1: Type: text/plain, Size: 4953 bytes --]
Some users of rhashtables might need to move an object from one table
to another - this appears to be the reason for the incomplete usage
of NULLS markers.
To support these, we store a unique NULLS_MARKER at the end of
each chain, and when a search fails to find a match, we check
if the NULLS marker found was the expected one. If not, the search
may not have examined all objects in the target bucket, so it is
repeated.
The unique NULLS_MARKER is derived from the address of the
head of the chain. As this cannot be derived at load-time the
static rhnull in rht_bucket_nested() needs to be initialised
at run time.
Any caller of a lookup function must still be prepared for the
possibility that the object returned is in a different table - it
might have been there for some time.
Note that this does NOT provide support for other uses of
NULLS_MARKERs such as allocating with SLAB_TYPESAFE_BY_RCU or changing
the key of an object and re-inserting it in the same table.
These could only be done safely if new objects were inserted
at the *start* of a hash chain, and that is not currently the case.
Signed-off-by: NeilBrown <neilb@suse.com>
---
I sent this patch back in July, but I made a bit of a mess of it.
Here it is again, with a bit more care.
Previously in
Commit: 9f9a707738aa ("rhashtable: remove nulls_base and related code.")
I removed some 'nulls' related code that wasn't being used and wasn't
usable.
This patch adds code to provide the functionality that the removed
code was intended for.
Specifically, it is now possible to move an object in one rhashtable
into another rhashtable, and still provide correct lookup(etc)
semantics.
If a lookup is performed on a table from which objects can be moved,
then a found object will have been in the table at some point during the
looku, but might not still be there. If that lookup returns failure,
then you can be sure that there was no matching object in the table for
the entire duration of the lookup.
This patch doesn't provide code to safely move an object. It just adds
enough code that the already-existing references to NULLS_MARKER()s make
sense.
NeilBrown
include/linux/rhashtable.h | 25 +++++++++++++++++--------
lib/rhashtable.c | 8 +++++---
2 files changed, 22 insertions(+), 11 deletions(-)
diff --git a/include/linux/rhashtable.h b/include/linux/rhashtable.h
index eb7111039247..8cc240f14834 100644
--- a/include/linux/rhashtable.h
+++ b/include/linux/rhashtable.h
@@ -75,8 +75,10 @@ struct bucket_table {
struct rhash_head __rcu *buckets[] ____cacheline_aligned_in_smp;
};
+#define RHT_NULLS_MARKER(ptr) \
+ ((void *)NULLS_MARKER(((unsigned long) (ptr)) >> 1))
#define INIT_RHT_NULLS_HEAD(ptr) \
- ((ptr) = (typeof(ptr)) NULLS_MARKER(0))
+ ((ptr) = RHT_NULLS_MARKER(&(ptr)))
static inline bool rht_is_a_nulls(const struct rhash_head *ptr)
{
@@ -471,6 +473,7 @@ static inline struct rhash_head *__rhashtable_lookup(
.ht = ht,
.key = key,
};
+ struct rhash_head __rcu * const *head;
struct bucket_table *tbl;
struct rhash_head *he;
unsigned int hash;
@@ -478,13 +481,19 @@ static inline struct rhash_head *__rhashtable_lookup(
tbl = rht_dereference_rcu(ht->tbl, ht);
restart:
hash = rht_key_hashfn(ht, tbl, key, params);
- rht_for_each_rcu(he, tbl, hash) {
- if (params.obj_cmpfn ?
- params.obj_cmpfn(&arg, rht_obj(ht, he)) :
- rhashtable_compare(&arg, rht_obj(ht, he)))
- continue;
- return he;
- }
+ head = rht_bucket(tbl, hash);
+ do {
+ rht_for_each_rcu_continue(he, *head, tbl, hash) {
+ if (params.obj_cmpfn ?
+ params.obj_cmpfn(&arg, rht_obj(ht, he)) :
+ rhashtable_compare(&arg, rht_obj(ht, he)))
+ continue;
+ return he;
+ }
+ /* An object might have been moved to a different hash chain,
+ * while we walk along it - better check and retry.
+ */
+ } while (he != RHT_NULLS_MARKER(head));
/* Ensure we see any new tables. */
smp_rmb();
diff --git a/lib/rhashtable.c b/lib/rhashtable.c
index 30526afa8343..852ffa5160f1 100644
--- a/lib/rhashtable.c
+++ b/lib/rhashtable.c
@@ -1179,8 +1179,7 @@ struct rhash_head __rcu **rht_bucket_nested(const struct bucket_table *tbl,
unsigned int hash)
{
const unsigned int shift = PAGE_SHIFT - ilog2(sizeof(void *));
- static struct rhash_head __rcu *rhnull =
- (struct rhash_head __rcu *)NULLS_MARKER(0);
+ static struct rhash_head __rcu *rhnull;
unsigned int index = hash & ((1 << tbl->nest) - 1);
unsigned int size = tbl->size >> tbl->nest;
unsigned int subhash = hash;
@@ -1198,8 +1197,11 @@ struct rhash_head __rcu **rht_bucket_nested(const struct bucket_table *tbl,
subhash >>= shift;
}
- if (!ntbl)
+ if (!ntbl) {
+ if (!rhnull)
+ INIT_RHT_NULLS_HEAD(rhnull);
return &rhnull;
+ }
return &ntbl[subhash].bucket;
--
2.14.0.rc0.dirty
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]
^ permalink raw reply related
* [net:master 25/27] drivers/net/ethernet/qlogic/qed/qed_rdma.h:186:79: error: expected ';' before '}' token
From: kbuild test robot @ 2018-11-14 13:16 UTC (permalink / raw)
To: Michal Kalderon; +Cc: kbuild-all, netdev, Denis Bolotin
[-- Attachment #1: Type: text/plain, Size: 1849 bytes --]
tree: https://git.kernel.org/pub/scm/linux/kernel/git/davem/net.git master
head: db8ddde766adf09605b5282e7978fa0ba76c3ee3
commit: 291d57f67d2449737d1e370ab5b9a583818eaa0c [25/27] qed: Fix rdma_info structure allocation
config: i386-allmodconfig (attached as .config)
compiler: gcc-7 (Debian 7.3.0-1) 7.3.0
reproduce:
git checkout 291d57f67d2449737d1e370ab5b9a583818eaa0c
# save the attached .config to linux build tree
make ARCH=i386
All errors (new ones prefixed by >>):
In file included from drivers/net/ethernet/qlogic/qed/qed_cxt.c:49:0:
drivers/net/ethernet/qlogic/qed/qed_rdma.h: In function 'qed_rdma_info_alloc':
>> drivers/net/ethernet/qlogic/qed/qed_rdma.h:186:79: error: expected ';' before '}' token
static inline int qed_rdma_info_alloc(struct qed_hwfn *p_hwfn) {return -EINVAL}
^
vim +186 drivers/net/ethernet/qlogic/qed/qed_rdma.h
176
177 #if IS_ENABLED(CONFIG_QED_RDMA)
178 void qed_rdma_dpm_bar(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt);
179 void qed_rdma_dpm_conf(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt);
180 int qed_rdma_info_alloc(struct qed_hwfn *p_hwfn);
181 void qed_rdma_info_free(struct qed_hwfn *p_hwfn);
182 #else
183 static inline void qed_rdma_dpm_conf(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt) {}
184 static inline void qed_rdma_dpm_bar(struct qed_hwfn *p_hwfn,
185 struct qed_ptt *p_ptt) {}
> 186 static inline int qed_rdma_info_alloc(struct qed_hwfn *p_hwfn) {return -EINVAL}
187 static inline void qed_rdma_info_free(struct qed_hwfn *p_hwfn) {}
188 #endif
189
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 66006 bytes --]
^ permalink raw reply
* [PATCH AUTOSEL 4.4 5/8] netfilter: xt_IDLETIMER: add sysfs filename checking routine
From: Sasha Levin @ 2018-11-14 22:26 UTC (permalink / raw)
To: stable, linux-kernel
Cc: Taehee Yoo, Pablo Neira Ayuso, Sasha Levin, netfilter-devel,
coreteam, netdev
In-Reply-To: <20181114222618.100378-1-sashal@kernel.org>
From: Taehee Yoo <ap420073@gmail.com>
[ Upstream commit 54451f60c8fa061af9051a53be9786393947367c ]
When IDLETIMER rule is added, sysfs file is created under
/sys/class/xt_idletimer/timers/
But some label name shouldn't be used.
".", "..", "power", "uevent", "subsystem", etc...
So that sysfs filename checking routine is needed.
test commands:
%iptables -I INPUT -j IDLETIMER --timeout 1 --label "power"
splat looks like:
[95765.423132] sysfs: cannot create duplicate filename '/devices/virtual/xt_idletimer/timers/power'
[95765.433418] CPU: 0 PID: 8446 Comm: iptables Not tainted 4.19.0-rc6+ #20
[95765.449755] Call Trace:
[95765.449755] dump_stack+0xc9/0x16b
[95765.449755] ? show_regs_print_info+0x5/0x5
[95765.449755] sysfs_warn_dup+0x74/0x90
[95765.449755] sysfs_add_file_mode_ns+0x352/0x500
[95765.449755] sysfs_create_file_ns+0x179/0x270
[95765.449755] ? sysfs_add_file_mode_ns+0x500/0x500
[95765.449755] ? idletimer_tg_checkentry+0x3e5/0xb1b [xt_IDLETIMER]
[95765.449755] ? rcu_read_lock_sched_held+0x114/0x130
[95765.449755] ? __kmalloc_track_caller+0x211/0x2b0
[95765.449755] ? memcpy+0x34/0x50
[95765.449755] idletimer_tg_checkentry+0x4e2/0xb1b [xt_IDLETIMER]
[ ... ]
Fixes: 0902b469bd25 ("netfilter: xtables: idletimer target implementation")
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
net/netfilter/xt_IDLETIMER.c | 20 ++++++++++++++++++++
1 file changed, 20 insertions(+)
diff --git a/net/netfilter/xt_IDLETIMER.c b/net/netfilter/xt_IDLETIMER.c
index 1718f536689f..8a1d2af3eed0 100644
--- a/net/netfilter/xt_IDLETIMER.c
+++ b/net/netfilter/xt_IDLETIMER.c
@@ -116,6 +116,22 @@ static void idletimer_tg_expired(unsigned long data)
schedule_work(&timer->work);
}
+static int idletimer_check_sysfs_name(const char *name, unsigned int size)
+{
+ int ret;
+
+ ret = xt_check_proc_name(name, size);
+ if (ret < 0)
+ return ret;
+
+ if (!strcmp(name, "power") ||
+ !strcmp(name, "subsystem") ||
+ !strcmp(name, "uevent"))
+ return -EINVAL;
+
+ return 0;
+}
+
static int idletimer_tg_create(struct idletimer_tg_info *info)
{
int ret;
@@ -126,6 +142,10 @@ static int idletimer_tg_create(struct idletimer_tg_info *info)
goto out;
}
+ ret = idletimer_check_sysfs_name(info->label, sizeof(info->label));
+ if (ret < 0)
+ goto out_free_timer;
+
sysfs_attr_init(&info->timer->attr.attr);
info->timer->attr.attr.name = kstrdup(info->label, GFP_KERNEL);
if (!info->timer->attr.attr.name) {
--
2.17.1
^ permalink raw reply related
* [PATCH AUTOSEL 4.4 2/8] netfilter: ipset: actually allow allowable CIDR 0 in hash:net,port,net
From: Sasha Levin @ 2018-11-14 22:26 UTC (permalink / raw)
To: stable, linux-kernel
Cc: Eric Westbrook, Eric Westbrook, Jozsef Kadlecsik,
Pablo Neira Ayuso, Sasha Levin, netfilter-devel, coreteam, netdev
In-Reply-To: <20181114222618.100378-1-sashal@kernel.org>
From: Eric Westbrook <eric@westbrook.io>
[ Upstream commit 886503f34d63e681662057448819edb5b1057a97 ]
Allow /0 as advertised for hash:net,port,net sets.
For "hash:net,port,net", ipset(8) says that "either subnet
is permitted to be a /0 should you wish to match port
between all destinations."
Make that statement true.
Before:
# ipset create cidrzero hash:net,port,net
# ipset add cidrzero 0.0.0.0/0,12345,0.0.0.0/0
ipset v6.34: The value of the CIDR parameter of the IP address is invalid
# ipset create cidrzero6 hash:net,port,net family inet6
# ipset add cidrzero6 ::/0,12345,::/0
ipset v6.34: The value of the CIDR parameter of the IP address is invalid
After:
# ipset create cidrzero hash:net,port,net
# ipset add cidrzero 0.0.0.0/0,12345,0.0.0.0/0
# ipset test cidrzero 192.168.205.129,12345,172.16.205.129
192.168.205.129,tcp:12345,172.16.205.129 is in set cidrzero.
# ipset create cidrzero6 hash:net,port,net family inet6
# ipset add cidrzero6 ::/0,12345,::/0
# ipset test cidrzero6 fe80::1,12345,ff00::1
fe80::1,tcp:12345,ff00::1 is in set cidrzero6.
See also:
https://bugzilla.kernel.org/show_bug.cgi?id=200897
https://github.com/ewestbrook/linux/commit/df7ff6efb0934ab6acc11f003ff1a7580d6c1d9c
Signed-off-by: Eric Westbrook <linux@westbrook.io>
Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
net/netfilter/ipset/ip_set_hash_netportnet.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/net/netfilter/ipset/ip_set_hash_netportnet.c b/net/netfilter/ipset/ip_set_hash_netportnet.c
index 9a14c237830f..b259a5814965 100644
--- a/net/netfilter/ipset/ip_set_hash_netportnet.c
+++ b/net/netfilter/ipset/ip_set_hash_netportnet.c
@@ -213,13 +213,13 @@ hash_netportnet4_uadt(struct ip_set *set, struct nlattr *tb[],
if (tb[IPSET_ATTR_CIDR]) {
e.cidr[0] = nla_get_u8(tb[IPSET_ATTR_CIDR]);
- if (!e.cidr[0] || e.cidr[0] > HOST_MASK)
+ if (e.cidr[0] > HOST_MASK)
return -IPSET_ERR_INVALID_CIDR;
}
if (tb[IPSET_ATTR_CIDR2]) {
e.cidr[1] = nla_get_u8(tb[IPSET_ATTR_CIDR2]);
- if (!e.cidr[1] || e.cidr[1] > HOST_MASK)
+ if (e.cidr[1] > HOST_MASK)
return -IPSET_ERR_INVALID_CIDR;
}
@@ -492,13 +492,13 @@ hash_netportnet6_uadt(struct ip_set *set, struct nlattr *tb[],
if (tb[IPSET_ATTR_CIDR]) {
e.cidr[0] = nla_get_u8(tb[IPSET_ATTR_CIDR]);
- if (!e.cidr[0] || e.cidr[0] > HOST_MASK)
+ if (e.cidr[0] > HOST_MASK)
return -IPSET_ERR_INVALID_CIDR;
}
if (tb[IPSET_ATTR_CIDR2]) {
e.cidr[1] = nla_get_u8(tb[IPSET_ATTR_CIDR2]);
- if (!e.cidr[1] || e.cidr[1] > HOST_MASK)
+ if (e.cidr[1] > HOST_MASK)
return -IPSET_ERR_INVALID_CIDR;
}
--
2.17.1
^ 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