29 Jan, 2016
1 commit
-
When switchdev drivers process FDB notifications from the underlying
device they resolve the netdev to which the entry points to and notify
the bridge using the switchdev notifier.However, since the RTNL mutex is not held there is nothing preventing
the netdev from disappearing in the middle, which will cause
br_switchdev_event() to dereference a non-existing netdev.Make switchdev drivers hold the lock at the beginning of the
notification processing session and release it once it ends, after
notifying the bridge.Also, remove switchdev_mutex and fdb_lock, as they are no longer needed
when RTNL mutex is held.Fixes: 03bf0c281234 ("switchdev: introduce switchdev notifier")
Signed-off-by: Ido Schimmel
Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller
16 Jan, 2016
1 commit
-
After promisc mode management was introduced a bridge device could do
dev_set_promiscuity from its ndo_change_rx_flags() callback which in
turn can be called after the bridge's addr_list_lock has been taken
(e.g. by dev_uc_add). This causes a false positive lockdep splat because
the port interfaces' addr_list_lock is taken when br_manage_promisc()
runs after the bridge's addr list lock was already taken.
To remove the false positive introduce a custom bridge addr_list_lock
class and set it on bridge init.
A simple way to reproduce this is with the following:
$ brctl addbr br0
$ ip l add l br0 br0.100 type vlan id 100
$ ip l set br0 up
$ ip l set br0.100 up
$ echo 1 > /sys/class/net/br0/bridge/vlan_filtering
$ brctl addif br0 eth0
Splat:
[ 43.684325] =============================================
[ 43.684485] [ INFO: possible recursive locking detected ]
[ 43.684636] 4.4.0-rc8+ #54 Not tainted
[ 43.684755] ---------------------------------------------
[ 43.684906] brctl/1187 is trying to acquire lock:
[ 43.685047] (_xmit_ETHER){+.....}, at: [] dev_set_rx_mode+0x1e/0x40
[ 43.685460] but task is already holding lock:
[ 43.685618] (_xmit_ETHER){+.....}, at: [] dev_uc_add+0x27/0x80
[ 43.686015] other info that might help us debug this:
[ 43.686316] Possible unsafe locking scenario:[ 43.686743] CPU0
[ 43.686967] ----
[ 43.687197] lock(_xmit_ETHER);
[ 43.687544] lock(_xmit_ETHER);
[ 43.687886] *** DEADLOCK ***[ 43.688438] May be due to missing lock nesting notation
[ 43.688882] 2 locks held by brctl/1187:
[ 43.689134] #0: (rtnl_mutex){+.+.+.}, at: [] rtnl_lock+0x17/0x20
[ 43.689852] #1: (_xmit_ETHER){+.....}, at: [] dev_uc_add+0x27/0x80
[ 43.690575] stack backtrace:
[ 43.690970] CPU: 0 PID: 1187 Comm: brctl Not tainted 4.4.0-rc8+ #54
[ 43.691270] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.8.1-20150318_183358- 04/01/2014
[ 43.691770] ffffffff826a25c0 ffff8800369fb8e0 ffffffff81360ceb ffffffff826a25c0
[ 43.692425] ffff8800369fb9b8 ffffffff810d0466 ffff8800369fb968 ffffffff81537139
[ 43.693071] ffff88003a08c880 0000000000000000 00000000ffffffff 0000000002080020
[ 43.693709] Call Trace:
[ 43.693931] [] dump_stack+0x4b/0x70
[ 43.694199] [] __lock_acquire+0x1e46/0x1e90
[ 43.694483] [] ? netlink_broadcast_filtered+0x139/0x3e0
[ 43.694789] [] ? nlmsg_notify+0x5a/0xc0
[ 43.695064] [] lock_acquire+0xe5/0x1f0
[ 43.695340] [] ? dev_set_rx_mode+0x1e/0x40
[ 43.695623] [] _raw_spin_lock_bh+0x45/0x80
[ 43.695901] [] ? dev_set_rx_mode+0x1e/0x40
[ 43.696180] [] dev_set_rx_mode+0x1e/0x40
[ 43.696460] [] dev_set_promiscuity+0x3c/0x50
[ 43.696750] [] br_port_set_promisc+0x25/0x50 [bridge]
[ 43.697052] [] br_manage_promisc+0x8a/0xe0 [bridge]
[ 43.697348] [] br_dev_change_rx_flags+0x1e/0x20 [bridge]
[ 43.697655] [] __dev_set_promiscuity+0x132/0x1f0
[ 43.697943] [] __dev_set_rx_mode+0x82/0x90
[ 43.698223] [] dev_uc_add+0x5e/0x80
[ 43.698498] [] vlan_device_event+0x542/0x650 [8021q]
[ 43.698798] [] notifier_call_chain+0x5d/0x80
[ 43.699083] [] raw_notifier_call_chain+0x16/0x20
[ 43.699374] [] call_netdevice_notifiers_info+0x6e/0x80
[ 43.699678] [] call_netdevice_notifiers+0x16/0x20
[ 43.699973] [] br_add_if+0x47e/0x4c0 [bridge]
[ 43.700259] [] add_del_if+0x6e/0x80 [bridge]
[ 43.700548] [] br_dev_ioctl+0xaf/0xc0 [bridge]
[ 43.700836] [] dev_ifsioc+0x30c/0x3c0
[ 43.701106] [] dev_ioctl+0xf9/0x6f0
[ 43.701379] [] ? mntput_no_expire+0x5/0x450
[ 43.701665] [] ? mntput_no_expire+0xae/0x450
[ 43.701947] [] sock_do_ioctl+0x42/0x50
[ 43.702219] [] sock_ioctl+0x1e5/0x290
[ 43.702500] [] do_vfs_ioctl+0x2cb/0x5c0
[ 43.702771] [] SyS_ioctl+0x79/0x90
[ 43.703033] [] entry_SYSCALL_64_fastpath+0x16/0x7aCC: Vlad Yasevich
CC: Stephen Hemminger
CC: Bridge list
CC: Andy Gospodarek
CC: Roopa Prabhu
Fixes: 2796d0c648c9 ("bridge: Automatically manage port promiscuous mode.")
Reported-by: Andy Gospodarek
Signed-off-by: Nikolay Aleksandrov
Signed-off-by: David S. Miller
11 Jan, 2016
1 commit
-
Offload MDB changes per port to hardware
Signed-off-by: Elad Raz
Signed-off-by: Ido Schimmel
Signed-off-by: Jiri Pirko
Reviewed-by: Nikolay Aleksandrov
Signed-off-by: David S. Miller
09 Jan, 2016
1 commit
-
Pablo Neira Ayuso says:
====================
Netfilter updates for net-nextThe following patchset contains Netfilter updates for net-next, they are:
1) Release nf_tables objects on netns destructions via
nft_release_afinfo().2) Destroy basechain and rules on netdevice removal in the new netdev
family.3) Get rid of defensive check against removal of inactive objects in
nf_tables.4) Pass down netns pointer to our existing nfnetlink callbacks, as well
as commit() and abort() nfnetlink callbacks.5) Allow to invert limit expression in nf_tables, so we can throttle
overlimit traffic.6) Add packet duplication for the netdev family.
7) Add forward expression for the netdev family.
8) Define pr_fmt() in conntrack helpers.
9) Don't leave nfqueue configuration on inconsistent state in case of
errors, from Ken-ichirou MATSUZAWA, follow up patches are also from
him.10) Skip queue option handling after unbind.
11) Return error on unknown both in nfqueue and nflog command.
12) Autoload ctnetlink when NFQA_CFG_F_CONNTRACK is set.
13) Add new NFTA_SET_USERDATA attribute to store user data in sets,
from Carlos Falgueras.14) Add support for 64 bit byteordering changes nf_tables, from Florian
Westphal.15) Add conntrack byte/packet counter matching support to nf_tables,
also from Florian.
====================Signed-off-by: David S. Miller
07 Jan, 2016
4 commits
-
Notifying hardware about newly bridged port vlan-aware changes.
Signed-off-by: Elad Raz
Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller -
Notifying hardware about bridge vlan-aware changes.
Signed-off-by: Elad Raz
Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller -
Disallow adding interfaces to a bridge when vlan filtering operation
failed. Send the failure code to the user.Signed-off-by: Elad Raz
Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller
06 Jan, 2016
1 commit
-
[I stole this patch from Eric Biederman. He wrote:]
> There is no defined mechanism to pass network namespace information
> into /sbin/bridge-stp therefore don't even try to invoke it except
> for bridge devices in the initial network namespace.
>
> It is possible for unprivileged users to cause /sbin/bridge-stp to be
> invoked for any network device name which if /sbin/bridge-stp does not
> guard against unreasonable arguments or being invoked twice on the
> same network device could cause problems.[Hannes: changed patch using netns_eq]
Cc: Eric W. Biederman
Signed-off-by: Eric W. Biederman
Signed-off-by: Hannes Frederic Sowa
Signed-off-by: David S. Miller
01 Jan, 2016
1 commit
29 Dec, 2015
1 commit
-
We have to release the existing objects on netns removal otherwise we
leak them. Chains are unregistered in first place to make sure no
packets are walking on our rules and sets anymore.The object release happens by when we unregister the family via
nft_release_afinfo() which is called from nft_unregister_afinfo() from
the corresponding __net_exit path in every family.Reported-by: Patrick McHardy
Signed-off-by: Pablo Neira Ayuso
24 Dec, 2015
1 commit
-
kobj_to_dev has been defined in linux/device.h, so I replace to_dev
with it.Signed-off-by: Geliang Tang
Signed-off-by: David S. Miller
23 Dec, 2015
1 commit
-
The bridge's ageing time is offloaded to hardware when:
1) A port joins a bridge
2) The ageing time of the bridge is changedIn the first case the ageing time is offloaded as jiffies, but in the
second case it's offloaded as clock_t, which is what existing switchdev
drivers expect to receive.Fixes: 6ac311ae8bfb ("Adding switchdev ageing notification on port bridged")
Signed-off-by: Ido Schimmel
Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller
19 Dec, 2015
1 commit
-
Pablo Neira Ayuso says:
====================
Netfilter updates for net-nextThe following patchset contains the first batch of Netfilter updates for
the upcoming 4.5 kernel. This batch contains userspace netfilter header
compilation fixes, support for packet mangling in nf_tables, the new
tracing infrastructure for nf_tables and cgroup2 support for iptables.
More specifically, they are:1) Two patches to include dependencies in our netfilter userspace
headers to resolve compilation problems, from Mikko Rapeli.2) Four comestic cleanup patches for the ebtables codebase, from Ian Morris.
3) Remove duplicate include in the netfilter reject infrastructure,
from Stephen Hemminger.4) Two patches to simplify the netfilter defragmentation code for IPv6,
patch from Florian Westphal.5) Fix root ownership of /proc/net netfilter for unpriviledged net
namespaces, from Philip Whineray.6) Get rid of unused fields in struct nft_pktinfo, from Florian Westphal.
7) Add mangling support to our nf_tables payload expression, from
Patrick McHardy.8) Introduce a new netlink-based tracing infrastructure for nf_tables,
from Florian Westphal.9) Change setter functions in nfnetlink_log to be void, from
Rami Rosen.10) Add netns support to the cttimeout infrastructure.
11) Add cgroup2 support to iptables, from Tejun Heo.
12) Introduce nfnl_dereference_protected() in nfnetlink, from Florian.
13) Add support for mangling pkttype in the nf_tables meta expression,
also from Florian.BTW, I need that you pull net into net-next, I have another batch that
requires changes that I don't yet see in net.
====================Signed-off-by: David S. Miller
16 Dec, 2015
1 commit
-
switchdev drivers need to know the netdev on which the switchdev op was
invoked. For example, the STP state of a VLAN interface configured on top
of a port can change while being member in a bridge. In this case, the
underlying driver should only change the STP state of that particular
VLAN and not of all the VLANs configured on the port.However, current switchdev infrastructure only passes the port netdev down
to the driver. Solve that by passing the original device down to the
driver as part of the required switchdev object / attribute.This doesn't entail any change in current switchdev drivers. It simply
enables those supporting stacked devices to know the originating device
and act accordingly.Signed-off-by: Ido Schimmel
Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller
15 Dec, 2015
1 commit
-
Resolve conflict between commit 264640fc2c5f4f ("ipv6: distinguish frag
queues by device for multicast and link-local packets") from the net
tree and commit 029f7f3b8701c ("netfilter: ipv6: nf_defrag: avoid/free
clone operations") from the nf-next tree.Signed-off-by: Pablo Neira Ayuso
Conflicts:
net/ipv6/netfilter/nf_conntrack_reasm.c
09 Dec, 2015
1 commit
-
Only needed when meta nftrace rule(s) were added.
The assumption is that no such rules are active, so the call to
nft_trace_init is "never" needed.When nftrace rules are active, we always call the nft_trace_* functions,
but will only send netlink messages when all of the following are true:- traceinfo structure was initialised
- skb->nf_trace == 1
- at least one subscriber to trace group.Adding an extra conditional
(static_branch ... && skb->nf_trace)
nft_trace_init( ..)Is possible but results in a larger nft_do_chain footprint.
Signed-off-by: Florian Westphal
Acked-by: Patrick McHardy
Signed-off-by: Pablo Neira Ayuso
04 Dec, 2015
2 commits
-
Sometimes the drivers and other code would find it handy to know some
internal information about upper device being changed. So allow upper-code
to pass information down to notifier listeners during linking.Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller -
Eliminate netdev_master_upper_dev_link_private and pass priv directly as
a parameter of netdev_master_upper_dev_link.Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller
24 Nov, 2015
4 commits
-
Eliminate some checkpatch issues by improved layout of if statements.
No changes detected by objdiff.
Signed-off-by: Ian Morris
Signed-off-by: Pablo Neira Ayuso -
Change brace placement to eliminate checkpatch error.
No changes detected by objdiff.
Signed-off-by: Ian Morris
Signed-off-by: Pablo Neira Ayuso -
Changes comments to use netdev style.
No changes detected by objdiff.
Signed-off-by: Ian Morris
Signed-off-by: Pablo Neira Ayuso -
Fixes a bunch of issues detected by checkpatch with regards to code
indentation.No changes detected by objdiff.
Signed-off-by: Ian Morris
Signed-off-by: Pablo Neira Ayuso
17 Nov, 2015
1 commit
-
When NET_SWITCHDEV=n, switchdev_port_attr_set simply returns EOPNOTSUPP.
In this case we should not emit errors and warnings to the kernel log.Reported-by: Sander Eikelenboom
Tested-by: Christian Borntraeger
Fixes: 0bc05d585d38 ("switchdev: allow caller to explicitly request
attr_set as deferred")
Fixes: 6ac311ae8bfb ("Adding switchdev ageing notification on port
bridged")
Signed-off-by: Ido Schimmel
Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller
11 Nov, 2015
1 commit
-
This reverts commit 34c2d9fb0498c066afbe610b15e18995fd8be792.
There are 2 reasons for this revert:
1) The commit in question doesn't do what it says it does. The
description reads: "Allow bridge forward delay to be configured
when Spanning Tree is enabled." This was already the case before
the commit was made. What the commit actually do was disallow
invalid values or 'forward_delay' when STP was turned off.2) The above change was actually a change in the user observed
behavior and broke things like libvirt and other network configs
that set 'forward_delay' to 0 without enabling STP. The value
of 0 is actually used when STP is turned off to immediately mark
the bridge as forwarding.Signed-off-by: Vlad Yasevich
Signed-off-by: David S. Miller
03 Nov, 2015
3 commits
-
br_should_learn() is protected by RCU and not by RTNL, so use correct
flavor of nbp_vlan_group().Fixes: 907b1e6e83ed ("bridge: vlan: use proper rcu for the vlgrp
member")
Signed-off-by: Ido Schimmel
Acked-by: Nikolay Aleksandrov
Acked-by: Jiri Pirko
Signed-off-by: David S. Miller -
The flag used to indicate if a VLAN should be used for filtering - as
opposed to context only - on the bridge itself (e.g. br0) is called
'brentry' and not 'brvlan'.Signed-off-by: Ido Schimmel
Acked-by: Nikolay Aleksandrov
Signed-off-by: David S. Miller -
When adding a port to a bridge we initialize VLAN filtering on it. We do
not bail out in case an error occurred in nbp_vlan_init, as it can be
used as a non VLAN filtering bridge.However, if VLAN filtering is required and an error occurred in
nbp_vlan_init, we should set vlgrp to NULL, so that VLAN filtering
functions (e.g. br_vlan_find, br_get_pvid) will know the struct is
invalid and will not try to access it.Signed-off-by: Ido Schimmel
Signed-off-by: Nikolay Aleksandrov
Signed-off-by: David S. Miller
30 Oct, 2015
1 commit
-
Problem Description:
We can add fdbs pointing to the bridge with NULL ->dst but that has a
few race conditions because br_fdb_insert() is used which first creates
the fdb and then, after the fdb has been published/linked, sets
"is_local" to 1 and in that time frame if a packet arrives for that fdb
it may see it as non-local and either do a NULL ptr dereference in
br_forward() or attach the fdb to the port where it arrived, and later
br_fdb_insert() will make it local thus getting a wrong fdb entry.
Call chain br_handle_frame_finish() -> br_forward():
But in br_handle_frame_finish() in order to call br_forward() the dst
should not be local i.e. skb != NULL, whenever the dst is
found to be local skb is set to NULL so we can't forward it,
and here comes the problem since it's running only
with RCU when forwarding packets it can see the entry before "is_local"
is set to 1 and actually try to dereference NULL.
The main issue is that if someone sends a packet to the switch while
it's adding the entry which points to the bridge device, it may
dereference NULL ptr. This is needed now after we can add fdbs
pointing to the bridge. This poses a problem for
br_fdb_update() as well, while someone's adding a bridge fdb, but
before it has is_local == 1, it might get moved to a port if it comes
as a source mac and then it may get its "is_local" set to 1This patch changes fdb_create to take is_local and is_static as
arguments to set these values in the fdb entry before it is added to the
hash. Also adds null check for port in br_forward.Fixes: 3741873b4f73 ("bridge: allow adding of fdb entries pointing to the bridge device")
Reported-by: Nikolay Aleksandrov
Signed-off-by: Roopa Prabhu
Reviewed-by: Nikolay Aleksandrov
Acked-by: Stephen Hemminger
Signed-off-by: David S. Miller
22 Oct, 2015
1 commit
-
if_nlmsg_size() overestimates the minimum allocation size of netlink
dump request (when called from rtnl_calcit()) or the size of the
message (when called from rtnl_getlink()). This is because
ext_filter_mask is not supported by rtnl_link_get_af_size() and
rtnl_link_get_size().The over-estimation is significant when at least one netdev has many
VLANs configured (8 bytes for each configured VLAN).This patch-set "rightsizes" the protocol specific attribute size
calculation by propagating ext_filter_mask to rtnl_link_get_af_size()
and adding this a argument to get_link_af_size op in rtnl_af_ops.Bridge module already used filtering aware sizing for notifications.
br_get_link_af_size_filtered() is consistent with the modified
get_link_af_size op so it replaces br_get_link_af_size() in br_af_ops.
br_get_link_af_size() becomes unused and thus removed.Signed-off-by: Ronen Arad
Acked-by: Sridhar Samudrala
Signed-off-by: David S. Miller
21 Oct, 2015
1 commit
-
Configure ageing time to the HW for newly bridged device
CC: Scott Feldman
CC: Jiri Pirko
Signed-off-by: Elad Raz
Acked-by: Jiri Pirko
Acked-by: Scott Feldman
Signed-off-by: David S. Miller
17 Oct, 2015
2 commits
-
This merge resolves conflicts with 75aec9df3a78 ("bridge: Remove
br_nf_push_frag_xmit_sk") as part of Eric Biederman's effort to improve
netns support in the network stack that reached upstream via David's
net-next tree.Signed-off-by: Pablo Neira Ayuso
Conflicts:
net/bridge/br_netfilter_hooks.c -
since commit 8405a8fff3f8 ("netfilter: nf_qeueue: Drop queue entries on
nf_unregister_hook") all pending queued entries are discarded.So we can simply remove all of the owner handling -- when module is
removed it also needs to unregister all its hooks.Signed-off-by: Florian Westphal
Signed-off-by: Pablo Neira Ayuso
15 Oct, 2015
3 commits
-
Since spinlock is held here, defer the switchdev operation. Also, ensure
that defered switchdev ops are processed before port master device
is unlinked.Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller -
When object is used in deferred work, we cannot use pointers in
switchdev object structures because the memory they point at may be already
used by someone else. So rather do local copy of the value.Signed-off-by: Jiri Pirko
Acked-by: Scott Feldman
Reviewed-by: John Fastabend
Signed-off-by: David S. Miller -
Caller should know if he can call attr_set directly (when holding RTNL)
or if he has to defer the att_set processing for later.This also allows drivers to sleep inside attr_set and report operation
status back to switchdev core. Switchdev core then warns if status is
not ok, instead of silent errors happening in drivers.Benefit from newly introduced switchdev deferred ops infrastructure.
Signed-off-by: Jiri Pirko
Signed-off-by: David S. Miller
13 Oct, 2015
4 commits
-
Ido Schimmel reported a problem with switchdev devices because of the
order change of del_nbp operations, more specifically the move of
nbp_vlan_flush() which deletes all vlans and frees vlgrp after the
rx_handler has been unregistered. So in order to fix this move
vlan_flush back where it was and make it destroy the rhtable after
NULLing vlgrp and waiting a grace period to make sure noone can see it.Reported-by: Ido Schimmel
Signed-off-by: Nikolay Aleksandrov
Reviewed-by: Ido Schimmel
Signed-off-by: David S. Miller -
As Ido Schimmel pointed out the vlan_vid_del() code in nbp_vlan_flush is
unnecessary (and is actually a remnant of the old vlan code) so we can
remove it.Signed-off-by: Nikolay Aleksandrov
Reviewed-by: Ido Schimmel
Signed-off-by: David S. Miller -
br_fill_ifinfo is called by br_ifinfo_notify which can be called from
many contexts with different locks held, sometimes it relies upon
bridge's spinlock only which is a problem for the vlan code, so use
explicitly rcu for that to avoid problems.Signed-off-by: Nikolay Aleksandrov
Reviewed-by: Ido Schimmel
Signed-off-by: David S. Miller -
The bridge and port's vlgrp member is already used in RCU way, currently
we rely on the fact that it cannot disappear while the port exists but
that is error-prone and we might miss places with improper locking
(either RCU or RTNL must be held to walk the vlan_list). So make it
official and use RCU for vlgrp to catch offenders. Introduce proper vlgrp
accessors and use them consistently throughout the code.Signed-off-by: Nikolay Aleksandrov
Reviewed-by: Ido Schimmel
Signed-off-by: David S. Miller