Tonghao Zhang [Mon, 28 Jan 2019 23:28:05 +0000 (15:28 -0800)]
net/mlx5e: Update hw flows when encap source mac changed
When we offload tc filters to hardware, hardware flows can
be updated when mac of encap destination ip is changed.
But we ignore one case, that the mac of local encap ip can
be changed too, so we should also update them.
To fix it, add route_dev in mlx5e_encap_entry struct to save
the local encap netdevice, and when mac changed, kernel will
flush all the neighbour on the netdevice and send NETEVENT_NEIGH_UPDATE
event. The mlx5 driver will delete the flows and add them when neighbour
available again.
Fixes: 232c001398ae ("net/mlx5e: Add support to neighbour update flow") Cc: Hadar Hen Zion <hadarh@mellanox.com> Signed-off-by: Tonghao Zhang <xiangxia.m.yue@gmail.com> Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com> Acked-by: Saeed Mahameed <saeedm@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Manish Chopra [Wed, 6 Feb 2019 22:43:47 +0000 (14:43 -0800)]
qed*: Advance drivers version to 8.37.0.20
Version update for qed/qede modules.
Signed-off-by: Manish Chopra <manishc@marvell.com> Signed-off-by: Ariel Elior <aelior@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Rahul Verma [Wed, 6 Feb 2019 22:43:46 +0000 (14:43 -0800)]
qed: Change verbosity for coalescing message.
Fix unnecessary logging of message in an expected
default case where coalescing value read (via ethtool -c)
migh not be valid unless they are configured explicitly
in the hardware using ethtool -C.
Signed-off-by: Rahul Verma <rverma@marvell.com> Signed-off-by: Manish Chopra <manishc@marvell.com> Signed-off-by: Ariel Elior <aelior@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Under heavy traffic load, when changing number of channels via
ethtool (ethtool -L) which will cause interface to be reloaded,
it was observed that some packets gets transmitted on old TX
channel/queue id which doesn't really exist after the channel
configuration leads to system crash.
Add a safeguard in the driver by validating queue id through
ndo_select_queue() which is called before the ndo_start_xmit().
Signed-off-by: Sudarsana Reddy Kalluru <skalluru@marvell.com> Signed-off-by: Ariel Elior <aelior@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
qed: Consider TX tcs while deriving the max num_queues for PF.
Max supported queues is derived incorrectly in the case of multi-CoS.
Need to consider TCs while calculating num_queues for PF.
Signed-off-by: Sudarsana Reddy Kalluru <skalluru@marvell.com> Signed-off-by: Ariel Elior <aelior@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
qed: Assign UFP TC value to vlan priority in UFP mode.
In the case of Unified Fabric Port (UFP) mode, switch provides
the traffic class (TC) value to be used for the traffic.
Configure hardware to use this TC value for vlan priority.
Signed-off-by: Sudarsana Reddy Kalluru <skalluru@marvell.com> Signed-off-by: Ariel Elior <aelior@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Manish Chopra [Wed, 6 Feb 2019 22:43:42 +0000 (14:43 -0800)]
qed: Fix EQ full firmware assert.
When slowpath messages are sent with high rate, the resulting
events can lead to a FW assert in case they are not handled fast
enough (Event Queue Full assert). Attempt to send queued slowpath
messages only after the newly evacuated entries in the EQ ring
are indicated to FW.
Signed-off-by: Manish Chopra <manishc@marvell.com> Signed-off-by: Ariel Elior <aelior@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Dan Carpenter [Wed, 6 Feb 2019 15:35:15 +0000 (18:35 +0300)]
net: dsa: Fix NULL checking in dsa_slave_set_eee()
This function can't succeed if dp->pl is NULL. It will Oops inside the
call to return phylink_ethtool_get_eee(dp->pl, e);
Fixes: 1be52e97ed3e ("dsa: slave: eee: Allow ports to use phylink") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Vivien Didelot <vivien.didelot@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Russell King [Wed, 6 Feb 2019 10:54:54 +0000 (10:54 +0000)]
MAINTAINERS: add maintainer for SFF/SFP/SFP+ support
Add maintainer entry for SFF/SFP/SFP+ support.
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Mon, 4 Feb 2019 16:36:06 +0000 (08:36 -0800)]
rxrpc: bad unlock balance in rxrpc_recvmsg
When either "goto wait_interrupted;" or "goto wait_error;"
paths are taken, socket lock has already been released.
This patch fixes following syzbot splat :
WARNING: bad unlock balance detected!
5.0.0-rc4+ #59 Not tainted
-------------------------------------
syz-executor223/8256 is trying to release lock (sk_lock-AF_RXRPC) at:
[<ffffffff86651353>] rxrpc_recvmsg+0x6d3/0x3099 net/rxrpc/recvmsg.c:598
but there are no more locks to release!
other info that might help us debug this:
1 lock held by syz-executor223/8256:
#0: 00000000fa9ed0f4 (slock-AF_RXRPC){+...}, at: spin_lock_bh include/linux/spinlock.h:334 [inline]
#0: 00000000fa9ed0f4 (slock-AF_RXRPC){+...}, at: release_sock+0x20/0x1c0 net/core/sock.c:2798
Fixes: 248f219cb8bc ("rxrpc: Rewrite the data and ack handling code") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: David Howells <dhowells@redhat.com> Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Tue, 5 Feb 2019 23:38:44 +0000 (15:38 -0800)]
mISDN: fix a race in dev_expire_timer()
Since mISDN_close() uses dev->pending to iterate over active
timers, there is a chance that one timer got removed from the
->pending list in dev_expire_timer() but that the thread
has not called yet wake_up_interruptible()
So mISDN_close() could miss this and free dev before
completion of at least one dev_expire_timer()
syzbot was able to catch this race :
BUG: KASAN: use-after-free in register_lock_class+0x140c/0x1bf0 kernel/locking/lockdep.c:827
Write of size 8 at addr ffff88809fc18948 by task syz-executor1/24769
Memory state around the buggy address: ffff88809fc18800: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ffff88809fc18880: 00 fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
>ffff88809fc18900: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^ ffff88809fc18980: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc ffff88809fc18a00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Karsten Keil <isdn@linux-pingi.de> Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Andrew Lunn [Tue, 5 Feb 2019 23:02:58 +0000 (00:02 +0100)]
net: dsa: mv88e6xxx: Fix counting of ATU violations
The ATU port vector contains a bit per port of the switch. The code
wrongly used it as a port number, and incremented a port counter. This
resulted in the wrong interfaces counter being incremented, and
potentially going off the end of the array of ports.
Fix this by using the source port ID for the violation, which really
is a port number.
Reported-by: Chris Healy <Chris.Healy@zii.aero> Tested-by: Chris Healy <Chris.Healy@zii.aero> Fixes: 65f60e4582bd ("net: dsa: mv88e6xxx: Keep ATU/VTU violation statistics") Signed-off-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Or Gerlitz [Thu, 10 Jan 2019 18:37:36 +0000 (20:37 +0200)]
net/mlx5e: Properly set steering match levels for offloaded TC decap rules
The match level computed by the driver gets to be wrong for decap
rules with wildcarded inner packet match such as:
tc filter add dev vxlan_sys_4789 protocol all parent ffff: prio 2 flower
enc_dst_ip 192.168.0.9 enc_key_id 100 enc_dst_port 4789
action tunnel_key unset
action mirred egress redirect dev eth1
The FW errs for a missing matching meta-data indicator for the outer
headers (where we do have a match), and a wrong matching meta-data
indicator for the inner headers (where we don't have a match).
Fix that by taking into account the matching on the tunnel info and
relating the match level of the encapsulated packet to the firmware
inner headers indicator in case of decap.
As for vxlan we mandate a match on the tunnel udp dst port, and in general
we practically madndate a match on the source or dest ip for any IP tunnel,
the fix was done in a minimal manner around the tunnel match parsing code.
Fixes: d708f902989b ('net/mlx5e: Get the required HW match level while parsing TC flow matches') Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Reported-by: Slava Ovsiienko <viacheslavo@mellanox.com> Reviewed-by: Jianbo Liu <jianbol@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Raed Salem [Mon, 17 Dec 2018 09:40:06 +0000 (11:40 +0200)]
net/mlx5e: FPGA, fix Innova IPsec TX offload data path performance
At Innova IPsec TX offload data path a special software parser metadata
is used to pass some packet attributes to the hardware, this metadata
is passed using the Ethernet control segment of a WQE (a HW descriptor)
header.
The cited commit might nullify this header, hence the metadata is lost,
this caused a significant performance drop during hw offloading
operation.
Fix by restoring the metadata at the Ethernet control segment in case
it was nullified.
The following patchset contains Netfilter fixes for net:
1) Use CONFIG_NF_TABLES_INET from seltests, not NF_TABLES_INET.
From Naresh Kamboju.
2) Add a test to cover masquerading and redirect case, from Florian
Westphal.
3) Two packets coming from the same socket may race to set up NAT,
ending up with different tuples and the packet losing race being
dropped. Update nf_conntrack_tuple_taken() to exercise clash
resolution for this case. From Martynas Pumputis and Florian
Westphal.
4) Unbind anonymous sets from the commit and abort path, this fixes
a splat due to double set list removal/release in case that the
transaction needs to be aborted.
5) Do not preserve original output interface for packets that are
redirected in the output chain when ip6_route_me_harder() is
called. Otherwise packets end up going not going to the loopback
device. From Eli Cooper.
6) Fix bogus splat in nft_compat with CONFIG_REFCOUNT_FULL=y, this
also simplifies the existing logic to deal with the list insertions
of the xtables extensions. From Florian Westphal.
Diffstat look rather larger than usual because of the new selftest, but
Florian and I consider that having tests soon into the tree is good to
improve coverage. If there's a different policy in this regard, please,
let me know.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
netfilter: nft_compat: don't use refcount_inc on newly allocated entry
When I moved the refcount to refcount_t type I missed the fact that
refcount_inc() will result in use-after-free warning with
CONFIG_REFCOUNT_FULL=y builds.
The correct fix would be to init the reference count to 1 at allocation
time, but, unfortunately we cannot do this, as we can't undo that
in case something else fails later in the batch.
So only solution I see is to special-case the 'new entry' condition
and replace refcount_inc() with a "delayed" refcount_set(1) in this case,
as done here.
The .activate callback can be removed to simplify things, we only
need to make sure that deactivate() decrements/unlinks the entry
from the list at end of transaction phase (commit or abort).
Fixes: 12c44aba6618 ("netfilter: nft_compat: use refcnt_t type for nft_xt reference count") Reported-by: Jordan Glover <Golden_Miller83@protonmail.ch> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Eli Cooper [Mon, 21 Jan 2019 10:45:27 +0000 (18:45 +0800)]
netfilter: ipv6: Don't preserve original oif for loopback address
Commit 508b09046c0f ("netfilter: ipv6: Preserve link scope traffic
original oif") made ip6_route_me_harder() keep the original oif for
link-local and multicast packets. However, it also affected packets
for the loopback address because it used rt6_need_strict().
REDIRECT rules in the OUTPUT chain rewrite the destination to loopback
address; thus its oif should not be preserved. This commit fixes the bug
that redirected local packets are being dropped. Actually the packet was
not exactly dropped; Instead it was sent out to the original oif rather
than lo. When a packet with daddr ::1 is sent to the router, it is
effectively dropped.
Fixes: 508b09046c0f ("netfilter: ipv6: Preserve link scope traffic original oif") Signed-off-by: Eli Cooper <elicooper@gmx.com> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This looks fairly harmless (no actual deadlock occurs), and is
fixed in a similar way to c6894dec8ea9 ("bridge: fix lockdep
addr_list_lock false positive splat") by putting the addr_list_lock
in its own lockdep class.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Rundong Ge [Sat, 2 Feb 2019 14:29:35 +0000 (14:29 +0000)]
net: dsa: slave: Don't propagate flag changes on down slave interfaces
The unbalance of master's promiscuity or allmulti will happen after ifdown
and ifup a slave interface which is in a bridge.
When we ifdown a slave interface , both the 'dsa_slave_close' and
'dsa_slave_change_rx_flags' will clear the master's flags. The flags
of master will be decrease twice.
In the other hand, if we ifup the slave interface again, since the
slave's flags were cleared the 'dsa_slave_open' won't set the master's
flag, only 'dsa_slave_change_rx_flags' that triggered by 'br_add_if'
will set the master's flags. The flags of master is increase once.
Only propagating flag changes when a slave interface is up makes
sure this does not happen. The 'vlan_dev_change_rx_flags' had the
same problem and was fixed, and changes here follows that fix.
Fixes: 91da11f870f0 ("net: Distributed Switch Architecture protocol support") Signed-off-by: Rundong Ge <rdong.ge@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 4 Feb 2019 17:43:48 +0000 (09:43 -0800)]
Merge branch 's390-qeth-fixes'
Julian Wiedmann says:
====================
s390/qeth: fixes 2019-02-04
please apply the following four fixes to -net.
Patch 1 takes care of a common resource leak in various error paths, while the
second patch fixes a misordered kfree when cleaning up after an error.
The other two patches ensure that there's no stale work dangling on workqueues
when the qeth device has already been offlined and/or removed.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Julian Wiedmann [Mon, 4 Feb 2019 16:40:09 +0000 (17:40 +0100)]
s390/qeth: conclude all event processing before offlining a card
Work for Bridgeport events is currently placed on a driver-wide
workqueue. If the card is removed and freed while any such work is still
active, this causes a use-after-free.
So put the events on a per-card queue, where we can control their
lifetime. As we also don't want stale events to last beyond an
offline & online cycle, flush this queue when setting the card offline.
Fixes: b4d72c08b358 ("qeth: bridgeport support - basic control") Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Julian Wiedmann [Mon, 4 Feb 2019 16:40:08 +0000 (17:40 +0100)]
s390/qeth: cancel close_dev work before removing a card
A card's close_dev work is scheduled on a driver-wide workqueue. If the
card is removed and freed while the work is still active, this causes a
use-after-free.
So make sure that the work is completed before freeing the card.
Fixes: 0f54761d167f ("qeth: Support VEPA mode") Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Julian Wiedmann [Mon, 4 Feb 2019 16:40:07 +0000 (17:40 +0100)]
s390/qeth: fix use-after-free in error path
The error path in qeth_alloc_qdio_buffers() that takes care of
cleaning up the Output Queues is buggy. It first frees the queue, but
then calls qeth_clear_outq_buffers() with that very queue struct.
Make the call to qeth_clear_outq_buffers() part of the free action
(in the correct order), and while at it fix the naming of the helper.
Fixes: 0da9581ddb0f ("qeth: exploit asynchronous delivery of storage blocks") Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com> Reviewed-by: Alexandra Winter <wintera@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Julian Wiedmann [Mon, 4 Feb 2019 16:40:06 +0000 (17:40 +0100)]
s390/qeth: release cmd buffer in error paths
Whenever we fail before/while starting an IO, make sure to release the
IO buffer. Usually qeth_irq() would do this for us, but if the IO
doesn't even start we obviously won't get an interrupt for it either.
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Petr Machata [Mon, 4 Feb 2019 14:50:38 +0000 (14:50 +0000)]
net: cls_flower: Remove filter from mask before freeing it
In fl_change(), when adding a new rule (i.e. fold == NULL), a driver may
reject the new rule, for example due to resource exhaustion. By that
point, the new rule was already assigned a mask, and it was added to
that mask's hash table. The clean-up path that's invoked as a result of
the rejection however neglects to undo the hash table addition, and
proceeds to free the new rule, thus leaving a dangling pointer in the
hash table.
Fix by removing fnew from the mask's hash table before it is freed.
Fixes: 35cc3cefc4de ("net/sched: cls_flower: Reject duplicated rules also under skip_sw") Signed-off-by: Petr Machata <petrm@mellanox.com> Acked-by: Jiri Pirko <jiri@mellanox.com> Reviewed-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 4 Feb 2019 17:11:19 +0000 (09:11 -0800)]
Merge branch 'smc-fixes'
Ursula Braun says:
====================
net/smc: fixes 2019-02-04
here are more fixes in the smc code for the net tree:
Patch 1 fixes an IB-related problem with SMCR.
Patch 2 fixes a cursor problem for one-way traffic.
Patch 3 fixes a problem with RMB-reusage.
Patch 4 fixes a closing issue.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Ursula Braun [Mon, 4 Feb 2019 12:44:47 +0000 (13:44 +0100)]
net/smc: correct state change for peer closing
If some kind of closing is received from the peer while still in
state SMC_INIT, it means the peer has had an active connection and
closed the socket quickly before listen_work finished. This should
not result in a shortcut from state SMC_INIT to state SMC_CLOSED.
This patch adds the socket to the accept queue in state
SMC_APPCLOSEWAIT1. The socket reaches state SMC_CLOSED once being
accepted and closed with smc_release().
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ursula Braun [Mon, 4 Feb 2019 12:44:46 +0000 (13:44 +0100)]
net/smc: delete rkey first before switching to unused
Once RMBs are flagged as unused they are candidates for reuse.
Thus the LLC DELETE RKEY operaton should be made before flagging
the RMB as unused.
Fixes: c7674c001b11 ("net/smc: unregister rkeys of unused buffer") Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ursula Braun [Mon, 4 Feb 2019 12:44:45 +0000 (13:44 +0100)]
net/smc: fix sender_free computation
In some scenarios a separate consumer cursor update is necessary.
The decision is made in smc_tx_consumer_cursor_update(). The
sender_free computation could be wrong:
The rx confirmed cursor is always smaller than or equal to the
rx producer cursor. The parameters in the smc_curs_diff() call
have to be exchanged, otherwise sender_free might even be negative.
And if more data arrives local_rx_ctrl.prod might be updated, enabling
a cursor difference between local_rx_ctrl.prod and rx confirmed cursor
larger than the RMB size. This case is not covered by smc_curs_diff().
Thus function smc_curs_diff_large() is introduced here.
If a recvmsg() is processed in parallel, local_tx_ctrl.cons might
change during smc_cdc_msg_send. Make sure rx_curs_confirmed is updated
with the actually sent local_tx_ctrl.cons value.
Fixes: e82f2e31f559 ("net/smc: optimize consumer cursor updates") Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ursula Braun [Mon, 4 Feb 2019 12:44:44 +0000 (13:44 +0100)]
net/smc: preallocated memory for rdma work requests
The work requests for rdma writes are built in local variables within
function smc_tx_rdma_write(). This violates the rule that the work
request storage has to stay till the work request is confirmed by
a completion queue response.
This patch introduces preallocated memory for these work requests.
The storage is allocated, once a link (and thus a queue pair) is
established.
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
During sendmsg() a cloned skb is saved via dp83640_txtstamp() in
->tx_queue. After the NIC sends this packet, the PHY will reply with a
timestamp for that TX packet. If the cable is pulled at the right time I
don't see that packet. It might gets flushed as part of queue shutdown
on NIC's side.
Once the link is up again then after the next sendmsg() we enqueue
another skb in dp83640_txtstamp() and have two on the list. Then the PHY
will send a reply and decode_txts() attaches it to the first skb on the
list.
No crash occurs since refcounting works but we are one packet behind.
linuxptp/ptp4l usually closes the socket and opens a new one (in such a
timeout case) so those "stale" replies never get there. However it does
not resume normal operation anymore.
Purge old skbs in decode_txts().
Fixes: cb646e2b02b2 ("ptp: Added a clock driver for the National Semiconductor PHYTER.") Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de> Acked-by: Richard Cochran <richardcochran@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
netfilter: nf_tables: unbind set in rule from commit path
Anonymous sets that are bound to rules from the same transaction trigger
a kernel splat from the abort path due to double set list removal and
double free.
This patch updates the logic to search for the transaction that is
responsible for creating the set and disable the set list removal and
release, given the rule is now responsible for this. Lookup is reverse
since the transaction that adds the set is likely to be at the tail of
the list.
Moreover, this patch adds the unbind step to deliver the event from the
commit path. This should not be done from the worker thread, since we
have no guarantees of in-order delivery to the listener.
This patch removes the assumption that both activate and deactivate
callbacks need to be provided.
Fixes: cd5125d8f518 ("netfilter: nf_tables: split set destruction in deactivate and destroy phase") Reported-by: Mikhail Morfikov <mmorfikov@gmail.com> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
netfilter: nf_nat: skip nat clash resolution for same-origin entries
It is possible that two concurrent packets originating from the same
socket of a connection-less protocol (e.g. UDP) can end up having
different IP_CT_DIR_REPLY tuples which results in one of the packets
being dropped.
To illustrate this, consider the following simplified scenario:
1. Packet A and B are sent at the same time from two different threads
by same UDP socket. No matching conntrack entry exists yet.
Both packets cause allocation of a new conntrack entry.
2. get_unique_tuple gets called for A. No clashing entry found.
conntrack entry for A is added to main conntrack table.
3. get_unique_tuple is called for B and will find that the reply
tuple of B is already taken by A.
It will allocate a new UDP source port for B to resolve the clash.
4. conntrack entry for B cannot be added to main conntrack table
because its ORIGINAL direction is clashing with A and the REPLY
directions of A and B are not the same anymore due to UDP source
port reallocation done in step 3.
This patch modifies nf_conntrack_tuple_taken so it doesn't consider
colliding reply tuples if the IP_CT_DIR_ORIGINAL tuples are equal.
[ Florian: simplify patch to not use .allow_clash setting
and always ignore identical flows ]
Signed-off-by: Martynas Pumputis <martynas@weave.works> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Toshiaki Makita [Thu, 31 Jan 2019 11:40:30 +0000 (20:40 +0900)]
virtio_net: Account for tx bytes and packets on sending xdp_frames
Previously virtnet_xdp_xmit() did not account for device tx counters,
which caused confusions.
To be consistent with SKBs, account them on freeing xdp_frames.
Reported-by: David Ahern <dsahern@gmail.com> Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Xin Long [Sun, 3 Feb 2019 19:27:58 +0000 (03:27 +0800)]
sctp: check and update stream->out_curr when allocating stream_out
Now when using stream reconfig to add out streams, stream->out
will get re-allocated, and all old streams' information will
be copied to the new ones and the old ones will be freed.
So without stream->out_curr updated, next time when trying to
send from stream->out_curr stream, a panic would be caused.
This patch is to check and update stream->out_curr when
allocating stream_out.
v1->v2:
- define fa_index() to get elem index from stream->out_curr.
v2->v3:
- repost with no change.
Fixes: 5bbbbe32a431 ("sctp: introduce stream scheduler foundations") Reported-by: Ying Xu <yinxu@redhat.com> Reported-by: syzbot+e33a3a138267ca119c7d@syzkaller.appspotmail.com Signed-off-by: Xin Long <lucien.xin@gmail.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: systemport: Fix WoL with password after deep sleep
Broadcom STB chips support a deep sleep mode where all register
contents are lost. Because we were stashing the MagicPacket password
into some of these registers a suspend into that deep sleep then a
resumption would not lead to being able to wake-up from MagicPacket with
password again.
Fix this by keeping a software copy of the password and program it
during suspend.
Fixes: 83e82f4c706b ("net: systemport: add Wake-on-LAN support") Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 3 Feb 2019 19:06:25 +0000 (11:06 -0800)]
Merge branch 'vsock-virtio-hot-unplug'
Stefano Garzarella says:
====================
vsock/virtio: fix issues on device hot-unplug
These patches try to handle the hot-unplug of vsock virtio transport device in
a proper way.
Maybe move the vsock_core_init()/vsock_core_exit() functions in the module_init
and module_exit of vsock_virtio_transport module can't be the best way, but the
architecture of vsock_core forces us to this approach for now.
The vsock_core proto_ops expect a valid pointer to the transport device, so we
can't call vsock_core_exit() until there are open sockets.
v2 -> v3:
- Rebased on master
v1 -> v2:
- Fixed commit message of patch 1.
- Added Reviewed-by, Acked-by tags by Stefan
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
vsock/virtio: reset connected sockets on device removal
When the virtio transport device disappear, we should reset all
connected sockets in order to inform the users.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
vsock/virtio: fix kernel panic after device hot-unplug
virtio_vsock_remove() invokes the vsock_core_exit() also if there
are opened sockets for the AF_VSOCK protocol family. In this way
the vsock "transport" pointer is set to NULL, triggering the
kernel panic at the first socket activity.
This patch move the vsock_core_init()/vsock_core_exit() in the
virtio_vsock respectively in module_init and module_exit functions,
that cannot be invoked until there are open sockets.
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1609699 Reported-by: Yan Fu <yafu@redhat.com> Signed-off-by: Stefano Garzarella <sgarzare@redhat.com> Acked-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The original diagnosis was incorrect: it appears that the NIC had
PHY polling mode enabled, which meant that it overwrote the PHYs
advertisement register during negotiation.
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> Tested-by: Yonglong Liu <liuyonglong@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Cc: Gerrit Renker <gerrit@erg.abdn.ac.uk> Signed-off-by: David S. Miller <davem@davemloft.net>
Karsten Graul [Wed, 30 Jan 2019 17:51:08 +0000 (18:51 +0100)]
net/smc: fix use of variable in cleared area
Do not use pend->idx as index for the arrays because its value is
located in the cleared area. Use the existing local variable instead.
Without this fix the wrong area might be cleared.
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Karsten Graul [Wed, 30 Jan 2019 17:51:07 +0000 (18:51 +0100)]
net/smc: use device link provided in qp_context
The device field of the IB event structure does not always point to the
SMC IB device. Load the pointer from the qp_context which is always
provided to smc_ib_qp_event_handler() in the priv field. And for qp
events the affected port is given in the qp structure of the ibevent,
derive it from there.
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Karsten Graul [Wed, 30 Jan 2019 17:51:06 +0000 (18:51 +0100)]
net/smc: call smc_cdc_msg_send() under send_lock
Call smc_cdc_msg_send() under the connection send_lock to make sure all
send operations for one connection are serialized.
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Karsten Graul [Wed, 30 Jan 2019 17:51:05 +0000 (18:51 +0100)]
net/smc: do not wait under send_lock
smc_cdc_get_free_slot() might wait for free transfer buffers when using
SMC-R. This wait should not be done under the send_lock, which is a
spin_lock. This fixes a cpu loop in parallel threads waiting for the
send_lock.
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Karsten Graul [Wed, 30 Jan 2019 17:51:04 +0000 (18:51 +0100)]
net/smc: recvmsg and splice_read should return 0 after shutdown
When a socket was connected and is now shut down for read, return 0 to
indicate end of data in recvmsg and splice_read (like TCP) and do not
return ENOTCONN. This behavior is required by the socket api.
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Karsten Graul [Wed, 30 Jan 2019 17:51:03 +0000 (18:51 +0100)]
net/smc: don't wait for send buffer space when data was already sent
When there is no more send buffer space and at least 1 byte was already
sent then return to user space. The wait is only done when no data was
sent by the sendmsg() call.
This fixes smc_tx_sendmsg() which tried to always send all user data and
started to wait for free send buffer space when needed. During this wait
the user space program was blocked in the sendmsg() call and hence not
able to receive incoming data. When both sides were in such a situation
then the connection stalled forever.
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Karsten Graul [Wed, 30 Jan 2019 17:51:02 +0000 (18:51 +0100)]
net/smc: prevent races between smc_lgr_terminate() and smc_conn_free()
To prevent races between smc_lgr_terminate() and smc_conn_free() add an
extra check of the lgr field before accessing it, and cancel a delayed
free_work when a new smc connection is created.
This fixes the problem that free_work cleared the lgr variable but
smc_lgr_terminate() or smc_conn_free() still access it in parallel.
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Hans Wippel [Wed, 30 Jan 2019 17:51:01 +0000 (18:51 +0100)]
net/smc: allow 16 byte pnetids in netlink policy
Currently, users can only send pnetids with a maximum length of 15 bytes
over the SMC netlink interface although the maximum pnetid length is 16
bytes. This patch changes the SMC netlink policy to accept 16 byte
pnetids.
Signed-off-by: Hans Wippel <hwippel@linux.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ursula Braun [Wed, 30 Jan 2019 17:51:00 +0000 (18:51 +0100)]
net/smc: fix another sizeof to int comparison
Comparing an int to a size, which is unsigned, causes the int to become
unsigned, giving the wrong result. kernel_sendmsg can return a negative
error code.
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
In sctp_sendmesg(), when walking the list of endpoint associations, the
association can be dropped from the list, making the list corrupt.
Properly handle this by using list_for_each_entry_safe()
Fixes: 4910280503f3 ("sctp: add support for snd flag SCTP_SENDALL process in sendmsg") Reported-by: Secunia Research <vuln@secunia.com> Tested-by: Secunia Research <vuln@secunia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 1 Feb 2019 18:14:12 +0000 (10:14 -0800)]
Merge tag 'mac80211-for-davem-2019-02-01' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211
Johannes Berg says:
====================
Two more fixes:
* sometimes, not enough tailroom was allocated for
software-encrypted management frames in mac80211
* cfg80211 regulatory restore got an additional condition,
needs to rerun the checks after that condition changes
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Dan Carpenter [Fri, 1 Feb 2019 08:28:16 +0000 (11:28 +0300)]
skge: potential memory corruption in skge_get_regs()
The "p" buffer is 0x4000 bytes long. B3_RI_WTO_R1 is 0x190. The value
of "regs->len" is in the 1-0x4000 range. The bug here is that
"regs->len - B3_RI_WTO_R1" can be a negative value which would lead to
memory corruption and an abrupt crash.
Fixes: c3f8be961808 ("[PATCH] skge: expand ethtool debug register dump") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Johannes Berg [Fri, 1 Feb 2019 10:09:54 +0000 (11:09 +0100)]
cfg80211: call disconnect_wk when AP stops
Since we now prevent regulatory restore during STA disconnect
if concurrent AP interfaces are active, we need to reschedule
this check when the AP state changes. This fixes never doing
a restore when an AP is the last interface to stop. Or to put
it another way: we need to re-check after anything we check
here changes.
Cc: stable@vger.kernel.org Fixes: 113f3aaa81bd ("cfg80211: Prevent regulatory restore during STA disconnect in concurrent interfaces") Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Felix Fietkau [Tue, 29 Jan 2019 10:10:57 +0000 (11:10 +0100)]
mac80211: ensure that mgmt tx skbs have tailroom for encryption
Some drivers use IEEE80211_KEY_FLAG_SW_MGMT_TX to indicate that management
frames need to be software encrypted. Since normal data packets are still
encrypted by the hardware, crypto_tx_tailroom_needed_cnt gets decremented
after key upload to hw. This can lead to passing skbs to ccmp_encrypt_skb,
which don't have the necessary tailroom for software encryption.
Change the code to add tailroom for encrypted management packets, even if
crypto_tx_tailroom_needed_cnt is 0.
Cc: stable@vger.kernel.org Signed-off-by: Felix Fietkau <nbd@nbd.name> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Daniel Borkmann [Thu, 31 Jan 2019 22:18:22 +0000 (23:18 +0100)]
Merge branch 'bpf-lockdep-fixes'
Alexei Starovoitov says:
====================
v1->v2:
- reworded 2nd patch. It's a real dead lock. Not a false positive
- dropped the lockdep fix for up_read_non_owner in bpf_get_stackid
In addition to preempt_disable patch for socket filters
https://patchwork.ozlabs.org/patch/1032437/
First patch fixes lockdep false positive in percpu_freelist
Second patch fixes potential deadlock in bpf_prog_register
Third patch fixes another potential deadlock in stackmap access
from tracing bpf prog and from syscall.
====================
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
The map_lookup_elem used to not acquiring spinlock
in order to optimize the reader.
It was true until commit 557c0c6e7df8 ("bpf: convert stackmap to pre-allocation")
The syscall's map_lookup_elem(stackmap) calls bpf_stackmap_copy().
bpf_stackmap_copy() may find the elem no longer needed after the copy is done.
If that is the case, pcpu_freelist_push() saves this elem for reuse later.
This push requires a spinlock.
If a tracing bpf_prog got run in the middle of the syscall's
map_lookup_elem(stackmap) and this tracing bpf_prog is calling
bpf_get_stackid(stackmap) which also requires the same pcpu_freelist's
spinlock, it may end up with a dead lock situation as reported by
Eric Dumazet in https://patchwork.ozlabs.org/patch/1030266/
The situation is the same as the syscall's map_update_elem() which
needs to acquire the pcpu_freelist's spinlock and could race
with tracing bpf_prog. Hence, this patch fixes it by protecting
bpf_stackmap_copy() with this_cpu_inc(bpf_prog_active)
to prevent tracing bpf_prog from running.
A later syscall's map_lookup_elem commit f1a2e44a3aec ("bpf: add queue and stack maps")
also acquires a spinlock and races with tracing bpf_prog similarly.
Hence, this patch is forward looking and protects the majority
of the map lookups. bpf_map_offload_lookup_elem() is the exception
since it is for network bpf_prog only (i.e. never called by tracing
bpf_prog).
Fixes: 557c0c6e7df8 ("bpf: convert stackmap to pre-allocation") Reported-by: Eric Dumazet <eric.dumazet@gmail.com> Acked-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Since tracepoints_mutex will be taken in tracepoint_probe_register/unregister()
there is no need to take bpf_event_mutex too.
bpf_event_mutex is protecting modifications to prog array used in kprobe/perf bpf progs.
bpf_raw_tracepoints don't need to take this mutex.
Fixes: c4f6699dfcb8 ("bpf: introduce BPF_RAW_TRACEPOINT") Acked-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
It has been explained that is a false positive here:
https://lkml.org/lkml/2018/7/25/756
Recap:
- stackmap uses pcpu_freelist
- The lock in pcpu_freelist is a percpu lock
- stackmap is only used by tracing bpf_prog
- A tracing bpf_prog cannot be run if another bpf_prog
has already been running (ensured by the percpu bpf_prog_active counter).
Eric pointed out that this lockdep splats stops other
legit lockdep splats in selftests/bpf/test_progs.c.
Fix this by calling local_irq_save/restore for stackmap.
Another false positive had also been worked around by calling
local_irq_save in commit 89ad2fa3f043 ("bpf: fix lockdep splat").
That commit added unnecessary irq_save/restore to fast path of
bpf hash map. irqs are already disabled at that point, since htab
is holding per bucket spin_lock with irqsave.
Let's reduce overhead for htab by introducing __pcpu_freelist_push/pop
function w/o irqsave and convert pcpu_freelist_push/pop to irqsave
to be used elsewhere (right now only in stackmap).
It stops lockdep false positive in stackmap with a bit of acceptable overhead.
Fixes: 557c0c6e7df8 ("bpf: convert stackmap to pre-allocation") Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org> Reported-by: Eric Dumazet <eric.dumazet@gmail.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Disabled preemption is necessary for proper access to per-cpu maps
from BPF programs.
But the sender side of socket filters didn't have preemption disabled:
unix_dgram_sendmsg->sk_filter->sk_filter_trim_cap->bpf_prog_run_save_cb->BPF_PROG_RUN
and a combination of af_packet with tun device didn't disable either:
tpacket_snd->packet_direct_xmit->packet_pick_tx_queue->ndo_select_queue->
tun_select_queue->tun_ebpf_select_queue->bpf_prog_run_clear_cb->BPF_PROG_RUN
Disable preemption before executing BPF programs (both classic and extended).
Reported-by: Jann Horn <jannh@google.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Song Liu <songliubraving@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
bpf, selftests: fix handling of sparse CPU allocations
Previously, bpf_num_possible_cpus() had a bug when calculating a
number of possible CPUs in the case of sparse CPU allocations, as
it was considering only the first range or element of
/sys/devices/system/cpu/possible.
E.g. in the case of "0,2-3" (CPU 1 is not available), the function
returned 1 instead of 3.
This patch fixes the function by making it parse all CPU ranges and
elements.
Signed-off-by: Martynas Pumputis <m@lambda.lt> Acked-by: Yonghong Song <yhs@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Michael Chan [Thu, 31 Jan 2019 19:31:48 +0000 (14:31 -0500)]
bnxt_en: Disable interrupts when allocating CP rings or NQs.
When calling firmware to allocate a CP ring or NQ, an interrupt associated
with that ring may be generated immediately before the doorbell is even
setup after the firmware call returns. When servicing the interrupt, the
driver may crash when trying to access the doorbell.
Fix it by disabling interrupt on that vector until the doorbell is
set up.
Fixes: 697197e5a173 ("bnxt_en: Re-structure doorbells.") Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 31 Jan 2019 17:47:56 +0000 (09:47 -0800)]
Merge branch 'ieee802154-for-davem-2019-01-31' of git://git.kernel.org/pub/scm/linux/kernel/git/sschmidt/wpan
Stefan Schmidt says:
====================
pull-request: ieee802154 for net 2019-01-31
An update from ieee802154 for your *net* tree.
I waited a while to see if anything else comes up, but it seems this time
we only have one fixup patch for the -rc rounds.
Colin fixed some indentation in the mcr20a drivers. That's about it.
If there are any problems with taking these two before the final 5.0 let
me know.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Bart Van Assche [Wed, 30 Jan 2019 18:42:30 +0000 (10:42 -0800)]
lib/test_rhashtable: Make test_insert_dup() allocate its hash table dynamically
The test_insert_dup() function from lib/test_rhashtable.c passes a
pointer to a stack object to rhltable_init(). Allocate the hash table
dynamically to avoid that the following is reported with object
debugging enabled:
Cc: Thomas Graf <tgraf@suug.ch> Cc: Herbert Xu <herbert@gondor.apana.org.au> Cc: netdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Bart Van Assche <bvanassche@acm.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
Jacob Wen [Thu, 31 Jan 2019 07:18:56 +0000 (15:18 +0800)]
l2tp: copy 4 more bytes to linear part if necessary
The size of L2TPv2 header with all optional fields is 14 bytes.
l2tp_udp_recv_core only moves 10 bytes to the linear part of a
skb. This may lead to l2tp_recv_common read data outside of a skb.
This patch make sure that there is at least 14 bytes in the linear
part of a skb to meet the maximum need of l2tp_udp_recv_core and
l2tp_recv_common. The minimum size of both PPP HDLC-like frame and
Ethernet frame is larger than 14 bytes, so we are safe to do so.
Also remove L2TP_HDR_SIZE_NOSEQ, it is unused now.
Fixes: fd558d186df2 ("l2tp: Split pppol2tp patch into separate l2tp and ppp parts") Suggested-by: Guillaume Nault <gnault@redhat.com> Signed-off-by: Jacob Wen <jian.w.wen@oracle.com> Acked-by: Guillaume Nault <gnault@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Brian Norris [Tue, 29 Jan 2019 23:12:01 +0000 (15:12 -0800)]
ath10k: correct bus type for WCN3990
WCN3990 is SNOC, not PCI. This prevents probing WCN3990.
Fixes: 367c899f622c ("ath10k: add bus type check in ath10k_init_hw_params") Signed-off-by: Brian Norris <briannorris@chromium.org> Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Lorenzo Bianconi [Sun, 27 Jan 2019 11:15:49 +0000 (12:15 +0100)]
mt76x0: eeprom: fix chan_vs_power map in mt76x0_get_power_info
Report correct eeprom per channel power value.
Fix chan_vs_power map in mt76x0_get_power_info routine
Fixes: f2a2e819d672 ("mt76x0: remove eeprom dependency from mt76x0_get_power_info") Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Jose Abreu [Wed, 30 Jan 2019 14:54:21 +0000 (15:54 +0100)]
net: stmmac: Disable EEE mode earlier in XMIT callback
In stmmac xmit callback we use a different flow for TSO packets but TSO
xmit callback is not disabling the EEE mode.
Fix this by disabling earlier the EEE mode, i.e. before calling the TSO
xmit callback.
Signed-off-by: Jose Abreu <joabreu@synopsys.com> Cc: Joao Pinto <jpinto@synopsys.com> Cc: David S. Miller <davem@davemloft.net> Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com> Cc: Alexandre Torgue <alexandre.torgue@st.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jose Abreu [Wed, 30 Jan 2019 14:54:20 +0000 (15:54 +0100)]
net: stmmac: Send TSO packets always from Queue 0
The number of TSO enabled channels in HW can be different than the
number of total channels. There is no way to determined, at runtime, the
number of TSO capable channels and its safe to assume that if TSO is
enabled then at least channel 0 will be TSO capable.
Lets always send TSO packets from Queue 0.
Signed-off-by: Jose Abreu <joabreu@synopsys.com> Cc: Joao Pinto <jpinto@synopsys.com> Cc: David S. Miller <davem@davemloft.net> Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com> Cc: Alexandre Torgue <alexandre.torgue@st.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jose Abreu [Wed, 30 Jan 2019 14:54:19 +0000 (15:54 +0100)]
net: stmmac: Fallback to Platform Data clock in Watchdog conversion
If we don't have DT then stmmac_clk will not be available. Let's add a
new Platform Data field so that we can specify the refclk by this mean.
This way we can still use the coalesce command in PCI based setups.
Signed-off-by: Jose Abreu <joabreu@synopsys.com> Cc: Joao Pinto <jpinto@synopsys.com> Cc: David S. Miller <davem@davemloft.net> Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com> Cc: Alexandre Torgue <alexandre.torgue@st.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Wed, 30 Jan 2019 11:49:48 +0000 (12:49 +0100)]
ipvlan, l3mdev: fix broken l3s mode wrt local routes
While implementing ipvlan l3 and l3s mode for kubernetes CNI plugin,
I ran into the issue that while l3 mode is working fine, l3s mode
does not have any connectivity to kube-apiserver and hence all pods
end up in Error state as well. The ipvlan master device sits on
top of a bond device and hostns traffic to kube-apiserver (also running
in hostns) is DNATed from 10.152.183.1:443 to 139.178.29.207:37573
where the latter is the address of the bond0. While in l3 mode, a
curl to https://10.152.183.1:443 or to https://139.178.29.207:37573
works fine from hostns, neither of them do in case of l3s. In the
latter only a curl to https://127.0.0.1:37573 appeared to work where
for local addresses of bond0 I saw kernel suddenly starting to emit
ARP requests to query HW address of bond0 which remained unanswered
and neighbor entries in INCOMPLETE state. These ARP requests only
happen while in l3s.
Debugging this further, I found the issue is that l3s mode is piggy-
backing on l3 master device, and in this case local routes are using
l3mdev_master_dev_rcu(dev) instead of net->loopback_dev as per commit f5a0aab84b74 ("net: ipv4: dst for local input routes should use l3mdev
if relevant") and 5f02ce24c269 ("net: l3mdev: Allow the l3mdev to be
a loopback"). I found that reverting them back into using the
net->loopback_dev fixed ipvlan l3s connectivity and got everything
working for the CNI.
Now judging from 4fbae7d83c98 ("ipvlan: Introduce l3s mode") and the
l3mdev paper in [0] the only sole reason why ipvlan l3s is relying
on l3 master device is to get the l3mdev_ip_rcv() receive hook for
setting the dst entry of the input route without adding its own
ipvlan specific hacks into the receive path, however, any l3 domain
semantics beyond just that are breaking l3s operation. Note that
ipvlan also has the ability to dynamically switch its internal
operation from l3 to l3s for all ports via ipvlan_set_port_mode()
at runtime. In any case, l3 vs l3s soley distinguishes itself by
'de-confusing' netfilter through switching skb->dev to ipvlan slave
device late in NF_INET_LOCAL_IN before handing the skb to L4.
Minimal fix taken here is to add a IFF_L3MDEV_RX_HANDLER flag which,
if set from ipvlan setup, gets us only the wanted l3mdev_l3_rcv() hook
without any additional l3mdev semantics on top. This should also have
minimal impact since dev->priv_flags is already hot in cache. With
this set, l3s mode is working fine and I also get things like
masquerading pod traffic on the ipvlan master properly working.
Fixes: f5a0aab84b74 ("net: ipv4: dst for local input routes should use l3mdev if relevant") Fixes: 5f02ce24c269 ("net: l3mdev: Allow the l3mdev to be a loopback") Fixes: 4fbae7d83c98 ("ipvlan: Introduce l3s mode") Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Cc: Mahesh Bandewar <maheshb@google.com> Cc: David Ahern <dsa@cumulusnetworks.com> Cc: Florian Westphal <fw@strlen.de> Cc: Martynas Pumputis <m@lambda.lt> Acked-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jacob Wen [Wed, 30 Jan 2019 06:55:14 +0000 (14:55 +0800)]
l2tp: fix reading optional fields of L2TPv3
Use pskb_may_pull() to make sure the optional fields are in skb linear
parts, so we can safely read them later.
It's easy to reproduce the issue with a net driver that supports paged
skb data. Just create a L2TPv3 over IP tunnel and then generates some
network traffic.
Once reproduced, rx err in /sys/kernel/debug/l2tp/tunnels will increase.
Changes in v4:
1. s/l2tp_v3_pull_opt/l2tp_v3_ensure_opt_in_linear/
2. s/tunnel->version != L2TP_HDR_VER_2/tunnel->version == L2TP_HDR_VER_3/
3. Add 'Fixes' in commit messages.
Changes in v3:
1. To keep consistency, move the code out of l2tp_recv_common.
2. Use "net" instead of "net-next", since this is a bug fix.
Changes in v2:
1. Only fix L2TPv3 to make code simple.
To fix both L2TPv3 and L2TPv2, we'd better refactor l2tp_recv_common.
It's complicated to do so.
2. Reloading pointers after pskb_may_pull
Fixes: f7faffa3ff8e ("l2tp: Add L2TPv3 protocol support") Fixes: 0d76751fad77 ("l2tp: Add L2TPv3 IP encapsulation (no UDP) support") Fixes: a32e0eec7042 ("l2tp: introduce L2TPv3 IP encapsulation support for IPv6") Signed-off-by: Jacob Wen <jian.w.wen@oracle.com> Acked-by: Guillaume Nault <gnault@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
George Amanakis [Wed, 30 Jan 2019 03:50:13 +0000 (22:50 -0500)]
tun: move the call to tun_set_real_num_queues
Call tun_set_real_num_queues() after the increment of tun->numqueues
since the former depends on it. Otherwise, the number of queues is not
correctly accounted for, which results to warnings similar to:
"vnet0 selects TX queue 11, but real number of TX queues is 11".
Fixes: 0b7959b62573 ("tun: publish tfile after it's fully initialized") Reported-and-tested-by: George Amanakis <gamanakis@gmail.com> Signed-off-by: George Amanakis <gamanakis@gmail.com> Signed-off-by: Stanislav Fomichev <sdf@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Yohei Kanemaru [Tue, 29 Jan 2019 06:52:34 +0000 (15:52 +0900)]
ipv6: sr: clear IP6CB(skb) on SRH ip4ip6 encapsulation
skb->cb may contain data from previous layers (in an observed case
IPv4 with L3 Master Device). In the observed scenario, the data in
IPCB(skb)->frags was misinterpreted as IP6CB(skb)->frag_max_size,
eventually caused an unexpected IPv6 fragmentation in ip6_fragment()
through ip6_finish_output().
This patch clears IP6CB(skb), which potentially contains garbage data,
on the SRH ip4ip6 encapsulation.
Fixes: 32d99d0b6702 ("ipv6: sr: add support for ip4ip6 encapsulation") Signed-off-by: Yohei Kanemaru <yohei.kanemaru@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
====================
virtio_net: Fix problems around XDP tx and napi_tx
While I'm looking into how to account standard tx counters on XDP tx
processing, I found several bugs around XDP tx and napi_tx.
Patch1: Fix oops on error path. Patch2 depends on this.
Patch2: Fix memory corruption on freeing xdp_frames with napi_tx enabled.
Patch3: Minor fix patch5 depends on.
Patch4: Fix memory corruption on processing xdp_frames when XDP is disabled.
Also patch5 depends on this.
Patch5: Fix memory corruption on processing xdp_frames while XDP is being
disabled.
Patch6: Minor fix patch7 depends on.
Patch7: Fix memory corruption on freeing sk_buff or xdp_frames when a normal
queue is reused for XDP and vise versa.
v2:
- patch5: Make rcu_assign_pointer/synchronize_net conditional instead of
_virtnet_set_queues.
- patch7: Use napi_consume_skb() instead of dev_consume_skb_any()
====================
Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Signed-off-by: David S. Miller <davem@davemloft.net>
Toshiaki Makita [Tue, 29 Jan 2019 00:45:59 +0000 (09:45 +0900)]
virtio_net: Differentiate sk_buff and xdp_frame on freeing
We do not reset or free up unused buffers when enabling/disabling XDP,
so it can happen that xdp_frames are freed after disabling XDP or
sk_buffs are freed after enabling XDP on xdp tx queues.
Thus we need to handle both forms (xdp_frames and sk_buffs) regardless
of XDP setting.
One way to trigger this problem is to disable XDP when napi_tx is
enabled. In that case, virtnet_xdp_set() calls virtnet_napi_enable()
which kicks NAPI. The NAPI handler will call virtnet_poll_cleantx()
which invokes free_old_xmit_skbs() for queues which have been used by
XDP.
Note that even with this change we need to keep skipping
free_old_xmit_skbs() from NAPI handlers when XDP is enabled, because XDP
tx queues do not aquire queue locks.
- v2: Use napi_consume_skb() instead of dev_consume_skb_any()
Fixes: 4941d472bf95 ("virtio-net: do not reset during XDP set") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Toshiaki Makita [Tue, 29 Jan 2019 00:45:58 +0000 (09:45 +0900)]
virtio_net: Use xdp_return_frame to free xdp_frames on destroying vqs
put_page() can work as a fallback for freeing xdp_frames, but the
appropriate way is to use xdp_return_frame().
Fixes: cac320c850ef ("virtio_net: convert to use generic xdp_frame and xdp_return_frame API") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Jesper Dangaard Brouer <brouer@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Toshiaki Makita [Tue, 29 Jan 2019 00:45:57 +0000 (09:45 +0900)]
virtio_net: Don't process redirected XDP frames when XDP is disabled
Commit 8dcc5b0ab0ec ("virtio_net: fix ndo_xdp_xmit crash towards dev not
ready for XDP") tried to avoid access to unexpected sq while XDP is
disabled, but was not complete.
There was a small window which causes out of bounds sq access in
virtnet_xdp_xmit() while disabling XDP.
An example case of
- curr_queue_pairs = 6 (2 for SKB and 4 for XDP)
- online_cpu_num = xdp_queue_paris = 4
when XDP is enabled:
CPU 0 CPU 1
(Disabling XDP) (Processing redirected XDP frames)
virtnet_xdp_xmit()
virtnet_xdp_set()
_virtnet_set_queues()
set curr_queue_pairs (2)
check if rq->xdp_prog is not NULL
virtnet_xdp_sq(vi)
qp = curr_queue_pairs -
xdp_queue_pairs +
smp_processor_id()
= 2 - 4 + 1 = -1
sq = &vi->sq[qp] // out of bounds access
set xdp_queue_pairs (0)
rq->xdp_prog = NULL
Basically we should not change curr_queue_pairs and xdp_queue_pairs
while someone can read the values. Thus, when disabling XDP, assign NULL
to rq->xdp_prog first, and wait for RCU grace period, then change
xxx_queue_pairs.
Note that we need to keep the current order when enabling XDP though.
- v2: Make rcu_assign_pointer/synchronize_net conditional instead of
_virtnet_set_queues.
Fixes: 186b3c998c50 ("virtio-net: support XDP_REDIRECT") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Toshiaki Makita [Tue, 29 Jan 2019 00:45:56 +0000 (09:45 +0900)]
virtio_net: Fix out of bounds access of sq
When XDP is disabled, curr_queue_pairs + smp_processor_id() can be
larger than max_queue_pairs.
There is no guarantee that we have enough XDP send queues dedicated for
each cpu when XDP is disabled, so do not count drops on sq in that case.
Fixes: 5b8f3c8d30a6 ("virtio_net: Add XDP related stats") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Toshiaki Makita [Tue, 29 Jan 2019 00:45:55 +0000 (09:45 +0900)]
virtio_net: Fix not restoring real_num_rx_queues
When _virtnet_set_queues() failed we did not restore real_num_rx_queues.
Fix this by placing the change of real_num_rx_queues after
_virtnet_set_queues().
This order is also in line with virtnet_set_channels().
Fixes: 4941d472bf95 ("virtio-net: do not reset during XDP set") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Toshiaki Makita [Tue, 29 Jan 2019 00:45:54 +0000 (09:45 +0900)]
virtio_net: Don't call free_old_xmit_skbs for xdp_frames
When napi_tx is enabled, virtnet_poll_cleantx() called
free_old_xmit_skbs() even for xdp send queue.
This is bogus since the queue has xdp_frames, not sk_buffs, thus mangled
device tx bytes counters because skb->len is meaningless value, and even
triggered oops due to general protection fault on freeing them.
Since xdp send queues do not aquire locks, old xdp_frames should be
freed only in virtnet_xdp_xmit(), so just skip free_old_xmit_skbs() for
xdp send queues.
Similarly virtnet_poll_tx() called free_old_xmit_skbs(). This NAPI
handler is called even without calling start_xmit() because cb for tx is
by default enabled. Once the handler is called, it enabled the cb again,
and then the handler would be called again. We don't need this handler
for XDP, so don't enable cb as well as not calling free_old_xmit_skbs().
Also, we need to disable tx NAPI when disabling XDP, so
virtnet_poll_tx() can safely access curr_queue_pairs and
xdp_queue_pairs, which are not atomically updated while disabling XDP.
Fixes: b92f1e6751a6 ("virtio-net: transmit napi") Fixes: 7b0411ef4aa6 ("virtio-net: clean tx descriptors from rx napi") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Toshiaki Makita [Tue, 29 Jan 2019 00:45:53 +0000 (09:45 +0900)]
virtio_net: Don't enable NAPI when interface is down
Commit 4e09ff536284 ("virtio-net: disable NAPI only when enabled during
XDP set") tried to fix inappropriate NAPI enabling/disabling when
!netif_running(), but was not complete.
On error path virtio_net could enable NAPI even when !netif_running().
This can cause enabling NAPI twice on virtnet_open(), which would
trigger BUG_ON() in napi_enable().
Fixes: 4941d472bf95b ("virtio-net: do not reset during XDP set") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
====================
erspan: always reports output key to userspace
Erspan protocol relies on output key to set session id header field.
However TUNNEL_KEY bit is cleared in order to not add key field to
the external GRE header and so the configured o_key is not reported
to userspace.
Fix the issue adding TUNNEL_KEY bit to the o_flags parameter dumping
device info
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Lorenzo Bianconi [Mon, 28 Jan 2019 21:23:49 +0000 (22:23 +0100)]
net: ip6_gre: always reports o_key to userspace
As Erspan_v4, Erspan_v6 protocol relies on o_key to configure
session id header field. However TUNNEL_KEY bit is cleared in
ip6erspan_tunnel_xmit since ERSPAN protocol does not set the key field
of the external GRE header and so the configured o_key is not reported
to userspace. The issue can be triggered with the following reproducer:
$ip link add ip6erspan1 type ip6erspan local 2000::1 remote 2000::2 \
key 1 seq erspan_ver 1
$ip link set ip6erspan1 up
ip -d link sh ip6erspan1
ip6erspan1@NONE: <BROADCAST,MULTICAST> mtu 1422 qdisc noop state DOWN mode DEFAULT
link/ether ba:ff:09:24:c3:0e brd ff:ff:ff:ff:ff:ff promiscuity 0 minmtu 68 maxmtu 1500
ip6erspan remote 2000::2 local 2000::1 encaplimit 4 flowlabel 0x00000 ikey 0.0.0.1 iseq oseq
Fix the issue adding TUNNEL_KEY bit to the o_flags parameter in
ip6gre_fill_info
Fixes: 5a963eb61b7c ("ip6_gre: Add ERSPAN native tunnel support") Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Lorenzo Bianconi [Mon, 28 Jan 2019 21:23:48 +0000 (22:23 +0100)]
net: ip_gre: always reports o_key to userspace
Erspan protocol (version 1 and 2) relies on o_key to configure
session id header field. However TUNNEL_KEY bit is cleared in
erspan_xmit since ERSPAN protocol does not set the key field
of the external GRE header and so the configured o_key is not reported
to userspace. The issue can be triggered with the following reproducer:
$ip link add erspan1 type erspan local 192.168.0.1 remote 192.168.0.2 \
key 1 seq erspan_ver 1
$ip link set erspan1 up
$ip -d link sh erspan1
Fix the issue adding TUNNEL_KEY bit to the o_flags parameter in
ipgre_fill_info
Fixes: 84e54fe0a5ea ("gre: introduce native tunnel support for ERSPAN") Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mathias Thore [Mon, 28 Jan 2019 09:07:47 +0000 (10:07 +0100)]
ucc_geth: Reset BQL queue when stopping device
After a timeout event caused by for example a broadcast storm, when
the MAC and PHY are reset, the BQL TX queue needs to be reset as
well. Otherwise, the device will exhibit severe performance issues
even after the storm has ended.
Co-authored-by: David Gounaris <david.gounaris@infinera.com> Signed-off-by: Mathias Thore <mathias.thore@infinera.com> Signed-off-by: David S. Miller <davem@davemloft.net>