* [RFC PATCH 5/5] netlink: rx mmap: notify only when NL_MMAP_STATUS_VALID frame exists
@ 2015-07-22 13:23 Ken-ichirou MATSUZAWA
0 siblings, 0 replies; 2+ messages in thread
From: Ken-ichirou MATSUZAWA @ 2015-07-22 13:23 UTC (permalink / raw)
To: netdev
Signed-off-by: Ken-ichirou MATSUZAWA <chamas@h4.dion.ne.jp>
---
net/netlink/af_netlink.c | 28 ++++++++++++++++------------
1 file changed, 16 insertions(+), 12 deletions(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 7e1610e..8901acd 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -576,16 +576,6 @@ netlink_current_frame(const struct netlink_ring *ring,
return netlink_lookup_frame(ring, ring->head, status);
}
-static struct nl_mmap_hdr *
-netlink_previous_frame(const struct netlink_ring *ring,
- enum nl_mmap_status status)
-{
- unsigned int prev;
-
- prev = ring->head ? ring->head - 1 : ring->frame_max;
- return netlink_lookup_frame(ring, prev, status);
-}
-
static void netlink_increment_head(struct netlink_ring *ring)
{
ring->head = ring->head != ring->frame_max ? ring->head + 1 : 0;
@@ -606,6 +596,21 @@ static void netlink_forward_ring(struct netlink_ring *ring)
} while (ring->head != head);
}
+static bool netlink_has_valid_frame(struct netlink_ring *ring)
+{
+ unsigned int head = ring->head, pos = head;
+ const struct nl_mmap_hdr *hdr;
+
+ do {
+ hdr = __netlink_lookup_frame(ring, pos);
+ if (hdr->nm_status == NL_MMAP_STATUS_VALID)
+ return true;
+ pos = pos != ring->frame_max ? pos + 1 : 0;
+ } while (pos != head);
+
+ return false;
+}
+
static bool netlink_dump_space(struct netlink_sock *nlk)
{
struct netlink_ring *ring = &nlk->rx_ring;
@@ -653,8 +658,7 @@ static unsigned int netlink_poll(struct file *file, struct socket *sock,
spin_lock_bh(&sk->sk_receive_queue.lock);
if (nlk->rx_ring.pg_vec) {
- netlink_forward_ring(&nlk->rx_ring);
- if (!netlink_previous_frame(&nlk->rx_ring, NL_MMAP_STATUS_UNUSED))
+ if (netlink_has_valid_frame(&nlk->rx_ring))
mask |= POLLIN | POLLRDNORM;
}
spin_unlock_bh(&sk->sk_receive_queue.lock);
--
1.7.10.4
----- End forwarded message -----
^ permalink raw reply related [flat|nested] 2+ messages in thread* [RFC PATCH 0/5] netlink: mmap kernel panic and some issues
@ 2015-07-22 1:09 Ken-ichirou MATSUZAWA
2015-07-22 1:15 ` [RFC PATCH 5/5] netlink: rx mmap: notify only when NL_MMAP_STATUS_VALID frame exists Ken-ichirou MATSUZAWA
0 siblings, 1 reply; 2+ messages in thread
From: Ken-ichirou MATSUZAWA @ 2015-07-22 1:09 UTC (permalink / raw)
To: The netfilter developer mailinglist
Hello,
I got a kernel panic below when I dumped using mmaped netlink socket
while monitoring it by nlmon tap device. I realized it is because
mmaped netlink skb does not have skb_shared_info but don't know how
to fix it in sane. This patch series seems to work fine for me but
I'm not sure it's right or not.
Patch 1/5 added helper functions for mmaped netlink skb and applied
these at 2/5. I'm not sure I embed helper functions like this or add
skb functions and wrap it like alloc_skb_head() in
netlink_alloc_skb(). Patch 3/5 fixes nm_state for skb which is
allocated but not sent.
I noticed I can not send netlink message by using mmaped netlink
socket since:
commit: a8866ff6a5bce7d0ec465a63bc482a85c09b0d39
netlink: make the check for "send from tx_ring" deterministic
I found a msg->msg_iter.type was set to 1 (WRITE). It seems that we
need to accept it but reject KERNEL_DS. Patch 4/5 may fix it.
Talking about Patch 5/5, I receive many notifications which frame
status is NL_MMAP_STATUS_RESERVED from mmaped nflog poll() when I
specified QTHRESH or TIMEOUT nflog config option. This behavior
seems to be different from normal socket. And I don't need to be
notified that there is a frame I'm processing - SKIP in the ring
too.
It would be appreciate if someone consolidate patches or tell me how
to fix it.
Thanks,
[ 196.691844] Netfilter messages via NETLINK v0.30.
[ 196.742847] nf_conntrack version 0.5.0 (2943 buckets, 11772 max)
[ 196.787119] ctnetlink v0.93: registering with nfnetlink.
[ 211.177865] device eth1 entered promiscuous mode
[ 211.314466] bridge: automatic filtering via arp/ip/ip6tables has been deprecated. Update your scripts to load br_netfilter if you need this.
[ 211.319998] br0: port 1(eth1) entered forwarding state
[ 211.320419] br0: port 1(eth1) entered forwarding state
[ 211.466591] Ebtables v2.0 registered
[ 226.336171] br0: port 1(eth1) entered forwarding state
[ 300.957103] BUG: unable to handle kernel NULL pointer dereference at 0000000000000002
[ 300.958740] IP: [<ffffffff81482b48>] kfree_skb_list+0x18/0x30
[ 300.959814] PGD 177ae067 PUD 177c6067 PMD 0
[ 300.960958] Oops: 0000 [#1] SMP
[ 300.960958] Modules linked in: nlmon nf_conntrack_ipv4 nf_defrag_ipv4 ebt_redirect ebtable_broute ebtables x_tables bridge stp llc dummy nf_conntrack_netlink nf_conntrack nfnetlink netconsole binfmt_misc ttm drm_kms_helper drm ppdev snd_pcm snd_timer parport_pc snd parport soundcore acpi_cpufreq psmouse pcspkr i2c_piix4 evdev i2c_core processor button thermal_sys serio_raw configfs loop autofs4 ext4 crc16 mbcache jbd2 sg sr_mod cdrom ata_generic virtio_blk virtio_net ata_piix virtio_pci virtio_ring virtio libata scsi_mod floppy [last unloaded: netconsole]
[ 300.960958] CPU: 0 PID: 890 Comm: ulogd Not tainted 4.1.1 #3
[ 300.960958] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007
[ 300.960958] task: ffff8800129963d0 ti: ffff880017254000 task.ti: ffff880017254000
[ 300.960958] RIP: 0010:[<ffffffff81482b48>] [<ffffffff81482b48>] kfree_skb_list+0x18/0x30
[ 300.960958] RSP: 0018:ffff8800172577e8 EFLAGS: 00010202
[ 300.960958] RAX: 0000000000000000 RBX: ffff88001513c000 RCX: 000000005fb50000
[ 300.960958] RDX: 00000000ffffffff RSI: ffff88000012e000 RDI: 0000000000000002
[ 300.960958] RBP: ffff8800172577f8 R08: 0000000000000020 R09: 0000000000000578
[ 300.960958] R10: ffffffff818c4cc0 R11: 0000000000000000 R12: ffff88001747d800
[ 300.960958] R13: 0000000000000000 R14: 0000000000001000 R15: ffff8800157ed400
[ 300.960958] FS: 00007f92e6dc1700(0000) GS:ffff880017c00000(0000) knlGS:0000000000000000
[ 300.960958] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 300.960958] CR2: 0000000000000002 CR3: 0000000015100000 CR4: 00000000000006f0
[ 300.960958] Stack:
[ 300.960958] ffff880017666600 ffff88001513c000 ffff880017257828 ffffffff81482be5
[ 300.960958] ffff880017257828 ffff88001747d800 0000000000000000 ffff88000012e000
[ 300.960958] ffff880017257848 ffffffff81482cc6 ffff88001747d800 ffff88001747d800
[ 300.960958] Call Trace:
[ 300.960958] [<ffffffff81482be5>] ? skb_release_data+0x85/0xd0
[ 300.960958] [<ffffffff81482cc6>] ? __kfree_skb+0x16/0x90
[ 300.960958] [<ffffffffa033b16c>] ? nlmon_xmit+0x2c/0x30 [nlmon]
[ 300.960958] [<ffffffff81494043>] ? dev_hard_start_xmit+0x233/0x3e0
[ 300.960958] [<ffffffff8149442e>] ? netif_skb_features+0xfe/0x200
[ 300.960958] [<ffffffff81494770>] ? validate_xmit_skb+0x40/0x330
[ 300.960958] [<ffffffff81494f59>] ? __dev_queue_xmit+0x489/0x590
[ 300.960958] [<ffffffff814c2e26>] ? netlink_deliver_tap+0xe6/0x170
[ 300.960958] [<ffffffff814c2eeb>] ? __netlink_sendskb+0x3b/0x240
[ 300.960958] [<ffffffff814c57c6>] ? netlink_dump+0x1c6/0x2d0
[ 300.960958] [<ffffffff814c769a>] ? __netlink_dump_start+0x19a/0x1d0
[ 300.960958] [<ffffffffa02f4d20>] ? ctnetlink_get_conntrack+0xc0/0x25c [nf_conntrack_netlink]
[ 300.960958] [<ffffffffa02f2b20>] ? ctnetlink_dump_dying+0x20/0x20 [nf_conntrack_netlink]
[ 300.960958] [<ffffffffa02f0a40>] ? ctnetlink_nfqueue_attach_expect+0x170/0x170 [nf_conntrack_netlink]
[ 300.960958] [<ffffffff8131a15e>] ? __nla_reserve+0x4e/0x70
[ 300.960958] [<ffffffff8131a15e>] ? __nla_reserve+0x4e/0x70
[ 300.960958] [<ffffffffa02f4c60>] ? ctnetlink_nfqueue_parse+0x2e0/0x2e0 [nf_conntrack_netlink]
[ 300.960958] [<ffffffffa0056b7b>] ? nfnetlink_rcv_msg+0x28b/0x2a0 [nfnetlink]
[ 300.960958] [<ffffffff81494770>] ? validate_xmit_skb+0x40/0x330
[ 300.960958] [<ffffffffa00568f0>] ? nfnetlink_rcv+0xe0/0xe0 [nfnetlink]
[ 300.960958] [<ffffffff814c65d9>] ? netlink_rcv_skb+0xa9/0xd0
[ 300.960958] [<ffffffff814c6266>] ? netlink_unicast+0x126/0x1c0
[ 300.960958] [<ffffffff814c6ea6>] ? netlink_sendmsg+0x556/0x660
[ 300.960958] [<ffffffff8147770d>] ? sock_sendmsg+0x4d/0x60
[ 300.960958] [<ffffffff814791b4>] ? SYSC_sendto+0x104/0x180
[ 300.960958] [<ffffffff811d7eb9>] ? vfs_read+0xa9/0xe0
[ 300.960958] [<ffffffff811d87fc>] ? SyS_read+0x9c/0xd0
[ 300.960958] [<ffffffff81596bae>] ? system_call_fastpath+0x12/0x71
[ 300.960958] Code: 48 83 c4 08 5b c9 c3 66 66 66 2e 0f 1f 84 00 00 00 00 00 55 48 89 e5 53 48 83 ec 08 0f 1f 44 00 00 48 85 ff 74 15 0f 1f 44 00 00 <48> 8b 1f e8 f0 fc ff ff 48 85 db 48 89 df 75 f0 48 83 c4 08 5b
[ 300.960958] RIP [<ffffffff81482b48>] kfree_skb_list+0x18/0x30
[ 300.960958] RSP <ffff8800172577e8>
[ 300.960958] CR2: 0000000000000002
[ 300.960958] ---[ end trace fa655a8b26512358 ]---
[ 300.960958] Kernel panic - not syncing: Fatal exception in interrupt
[ 300.960958] Kernel Offset: disabled
[ 300.960958] ---[ end Kernel panic - not syncing: Fatal exception in interrupt
^ permalink raw reply [flat|nested] 2+ messages in thread* [RFC PATCH 5/5] netlink: rx mmap: notify only when NL_MMAP_STATUS_VALID frame exists
2015-07-22 1:09 [RFC PATCH 0/5] netlink: mmap kernel panic and some issues Ken-ichirou MATSUZAWA
@ 2015-07-22 1:15 ` Ken-ichirou MATSUZAWA
0 siblings, 0 replies; 2+ messages in thread
From: Ken-ichirou MATSUZAWA @ 2015-07-22 1:15 UTC (permalink / raw)
To: The netfilter developer mailinglist
Signed-off-by: Ken-ichirou MATSUZAWA <chamas@h4.dion.ne.jp>
---
net/netlink/af_netlink.c | 28 ++++++++++++++++------------
1 file changed, 16 insertions(+), 12 deletions(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 7e1610e..8901acd 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -576,16 +576,6 @@ netlink_current_frame(const struct netlink_ring *ring,
return netlink_lookup_frame(ring, ring->head, status);
}
-static struct nl_mmap_hdr *
-netlink_previous_frame(const struct netlink_ring *ring,
- enum nl_mmap_status status)
-{
- unsigned int prev;
-
- prev = ring->head ? ring->head - 1 : ring->frame_max;
- return netlink_lookup_frame(ring, prev, status);
-}
-
static void netlink_increment_head(struct netlink_ring *ring)
{
ring->head = ring->head != ring->frame_max ? ring->head + 1 : 0;
@@ -606,6 +596,21 @@ static void netlink_forward_ring(struct netlink_ring *ring)
} while (ring->head != head);
}
+static bool netlink_has_valid_frame(struct netlink_ring *ring)
+{
+ unsigned int head = ring->head, pos = head;
+ const struct nl_mmap_hdr *hdr;
+
+ do {
+ hdr = __netlink_lookup_frame(ring, pos);
+ if (hdr->nm_status == NL_MMAP_STATUS_VALID)
+ return true;
+ pos = pos != ring->frame_max ? pos + 1 : 0;
+ } while (pos != head);
+
+ return false;
+}
+
static bool netlink_dump_space(struct netlink_sock *nlk)
{
struct netlink_ring *ring = &nlk->rx_ring;
@@ -653,8 +658,7 @@ static unsigned int netlink_poll(struct file *file, struct socket *sock,
spin_lock_bh(&sk->sk_receive_queue.lock);
if (nlk->rx_ring.pg_vec) {
- netlink_forward_ring(&nlk->rx_ring);
- if (!netlink_previous_frame(&nlk->rx_ring, NL_MMAP_STATUS_UNUSED))
+ if (netlink_has_valid_frame(&nlk->rx_ring))
mask |= POLLIN | POLLRDNORM;
}
spin_unlock_bh(&sk->sk_receive_queue.lock);
--
1.7.10.4
^ permalink raw reply related [flat|nested] 2+ messages in thread
end of thread, other threads:[~2015-07-22 13:23 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-07-22 13:23 [RFC PATCH 5/5] netlink: rx mmap: notify only when NL_MMAP_STATUS_VALID frame exists Ken-ichirou MATSUZAWA
-- strict thread matches above, loose matches on Subject: below --
2015-07-22 1:09 [RFC PATCH 0/5] netlink: mmap kernel panic and some issues Ken-ichirou MATSUZAWA
2015-07-22 1:15 ` [RFC PATCH 5/5] netlink: rx mmap: notify only when NL_MMAP_STATUS_VALID frame exists Ken-ichirou MATSUZAWA
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.