Use IFF_NO_QUEUE to indicate that a device can run without a qdisc.
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently when a macvlan is being initialized and the lower device is
netif_carrier_ok(), the macvlan device doesn't run through
rfc2863_policy() and is left with UNKNOWN operstate. Fix it by adding an
unconditional linkwatch event for the new macvlan device. Similar fix is
already used by the 8021q device (see register_vlan_dev()). Also fix the
inconsistent state when the lower device has been down and its carrier
was changed (when a device is down NETDEV_CHANGE doesn't get generated).
The second issue can be seen f.e. when we have a macvlan on top of a 8021q
device which has been down and its real device has been changing carrier
states, after setting the 8021q device up, the macvlan device will have
the same carrier state as it was before even though the 8021q can now
have a different state.
Example for case 1:
4: eth2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast
state UP mode DEFAULT group default qlen 1000
$ ip l add l eth2 macvl0 type macvlan
$ ip l set macvl0 up
$ ip l sh macvl0
72: macvl0@eth2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc
noqueue state UNKNOWN mode DEFAULT group default
link/ether f6:0b:54:0a:9d:a3 brd ff:ff:ff:ff:ff:ff
Example for case 2 (order is important):
Prestate: eth2 UP/CARRIER, vlan1 down, vlan1-macvlan down
$ ip l set vlan1-macvlan up
$ ip l sh vlan1-macvlan
71: vlan1-macvlan@vlan1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500
qdisc noqueue state UNKNOWN mode DEFAULT group default
link/ether 4a:b8:44:56:b9:b9 brd ff:ff:ff:ff:ff:ff
[ eth2 loses CARRIER before vlan1 has been UP-ed ]
$ ip l sh eth2
4: eth2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast
state DOWN mode DEFAULT group default qlen 1000
link/ether 52:54:00:bf:57:16 brd ff:ff:ff:ff:ff:ff
$ ip l sh vlan1-macvlan
71: vlan1-macvlan@vlan1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500
qdisc noqueue state UNKNOWN mode DEFAULT group default
link/ether 4a:b8:44:56:b9:b9 brd ff:ff:ff:ff:ff:ff
$ ip l set vlan1 up
$ ip l sh vlan1
70: vlan1@eth2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc
noqueue state LOWERLAYERDOWN mode DEFAULT group default qlen 1000
link/ether 52:54:00:bf:57:16 brd ff:ff:ff:ff:ff:ff
$ ip l sh vlan1-macvlan
71: vlan1-macvlan@vlan1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500
qdisc noqueue state UNKNOWN mode DEFAULT group default
link/ether 4a:b8:44:56:b9:b9 brd ff:ff:ff:ff:ff:ff
vlan1-macvlan is still UP, still has carrier and is still in the same
operstate as before. After the patch in case 1 macvl0 has state UP as it
should and in case 2 vlan1-macvlan has state LOWERLAYERDOWN again as it
should. Note that while the lower macvlan device is down their carrier
and thus operstate can go out of sync but that will be fixed once the
lower device goes up again.
This behaviour seems to have been present since beginning of git history.
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
These netif flags are unnecessary convolutions. It is more
straightforward to just use NETIF_F_HW_CSUM, NETIF_F_IP_CSUM,
and NETIF_F_IPV6_CSUM directly.
This patch also:
- Cleans up can_checksum_protocol
- Simplifies netdev_intersect_features
Signed-off-by: Tom Herbert <tom@herbertland.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The name NETIF_F_ALL_CSUM is a misnomer. This does not correspond to the
set of features for offloading all checksums. This is a mask of the
checksum offload related features bits. It is incorrect to set both
NETIF_F_HW_CSUM and NETIF_F_IP_CSUM or NETIF_F_IPV6 at the same time for
features of a device.
This patch:
- Changes instances of NETIF_F_ALL_CSUM to NETIF_F_CSUM_MASK (where
NETIF_F_ALL_CSUM is being used as a mask).
- Changes bonding, sfc/efx, ipvlan, macvlan, vlan, and team drivers to
use NEITF_F_HW_CSUM in features list instead of NETIF_F_ALL_CSUM.
Signed-off-by: Tom Herbert <tom@herbertland.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Reset pskb in macvlan_handle_frame in case skb_share_check returned a
clone.
Fixes: 8a4eb5734e ("net: introduce rx_handler results and logic around that")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function ip_defrag is called on both the input and the output
paths of the networking stack. In particular conntrack when it is
tracking outbound packets from the local machine calls ip_defrag.
So add a struct net parameter and stop making ip_defrag guess which
network namespace it needs to defragment packets in.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Macvlan/macvtap devices don't need to segment multiple tagged packets
since the lower devices can segment them.
Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a macvlan device is placed in promiscuous mode, it currently
just sets it's multicast mask to permissive, but doesn't change
the state of the lower device. As a result, not all multicast
traffic can be received on such device. Additionally, none of
a vlan traffic can be received on such device as well.
This patch propagates the promiscuous mode setting to lower device
so that lower device may receive all packets that macvlan may
be interested in.
Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Don't use dev->iflink anymore.
CC: Patrick McHardy <kaber@trash.net>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now that there are no more users kill dev_rebuild_header and all of it's
implementations.
This is long overdue.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Assign rtnl_link_ops->get_link_net() callback so that IFLA_LINK_NETNSID is
added to rtnetlink messages.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If device is already used as an ipvlan port then refuse to
use it as a macvlan port at early stage of port creation.
thost1:~# ip link add link eth0 ipvl0 type ipvlan
thost1:~# echo $?
0
thost1:~# ip link add link eth0 mvl0 type macvlan
RTNETLINK answers: Device or resource busy
thost1:~# echo $?
2
thost1:~#
Signed-off-by: Mahesh Bandewar <maheshb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit fbe168ba91 ("net: generic dev_disable_lro() stacked
device handling"), dev_disable_lro() zeroes NETIF_F_LRO feature flag
first for a macvlan device and then for its lower device. As an attempt
to set NETIF_F_LRO to zero is ignored, dev_disable_lro() issues a
warning and taints kernel.
Allowing NETIF_F_LRO to be set independently of the lower device
consists of three parts:
- add the flag to hw_features to allow toggling it
- allow setting it to 0 even if lower device has the flag set
- add the flag to MACVLAN_FEATURES to restore copying from lower
device on macvlan creation
Signed-off-by: Michal Kubecek <mkubecek@suse.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>
Do the work of parsing NDA_VLAN directly in rtnetlink code, pass simple
u16 vid to drivers from there.
Signed-off-by: Jiri Pirko <jiri@resnulli.us>
Acked-by: Andy Gospodarek <gospo@cumulusnetworks.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We do header check twice for a dodgy packet. One is done before
macvlan_start_xmit(), another is done before lower device's
ndo_start_xmit(). The first one seems redundant so this patch tries to
delay header check until a packet reaches its lower device (or macvtap)
through always enabling NETIF_F_GSO_ROBUST for macvlan device.
Cc: Patrick McHardy <kaber@trash.net>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We need to cancel the work queue after rcu grace period,
otherwise it can be rescheduled by incoming packets.
We need to purge queue if some skbs are still in it.
We can use __skb_queue_head_init() variant in
macvlan_process_broadcast()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: 412ca1550c ("macvlan: Move broadcasts into a work queue")
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
The netif_rx() call on the fast path of macvlan_handle_frame() appears to
be there to ensure that we properly throttle incoming packets. However, it
would appear as though the proper throttling is already in place for all
possible ingress paths, and that the call is redundant. If packets are arriving
from the physical NIC, we've already throttled them by this point. Otherwise,
if they are coming via macvlan_queue_xmit(), it calls either
'dev_forward_skb()', which ends up calling netif_rx_internal(), or else in
the broadcast case, we are throttling via macvlan_broadcast_enqueue().
The test results below are from off the box to an lxc instance running macvlan.
Once the tranactions/sec stop increasing, the cpu idle time has gone to 0.
Results are from a quad core Intel E3-1270 V2@3.50GHz box with bnx2x 10G card.
for i in {10,100,200,300,400,500};
do super_netperf $i -H $ip -t TCP_RR; done
Average of 5 runs.
trans/sec trans/sec
(3.17-rc7-net-next) (3.17-rc7-net-next + this patch)
---------- ----------
208101 211534 (+1.6%)
839493 850162 (+1.3%)
845071 844053 (-.12%)
816330 819623 (+.4%)
778700 789938 (+1.4%)
735984 754408 (+2.5%)
Signed-off-by: Jason Baron <jbaron@akamai.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pass last argument to macvlan_count_rx() as the correct bool type.
Signed-off-by: Jason Baron <jbaron@akamai.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Testing xmit_more support with netperf and connected UDP sockets,
I found strange dst refcount false sharing.
Current handling of IFF_XMIT_DST_RELEASE is not optimal.
Dropping dst in validate_xmit_skb() is certainly too late in case
packet was queued by cpu X but dequeued by cpu Y
The logical point to take care of drop/force is in __dev_queue_xmit()
before even taking qdisc lock.
As Julian Anastasov pointed out, need for skb_dst() might come from some
packet schedulers or classifiers.
This patch adds new helper to cleanly express needs of various drivers
or qdiscs/classifiers.
Drivers that need skb_dst() in their ndo_start_xmit() should call
following helper in their setup instead of the prior :
dev->priv_flags &= ~IFF_XMIT_DST_RELEASE;
->
netif_keep_dst(dev);
Instead of using a single bit, we use two bits, one being
eventually rebuilt in bonding/team drivers.
The other one, is permanent and blocks IFF_XMIT_DST_RELEASE being
rebuilt in bonding/team. Eventually, we could add something
smarter later.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Julian Anastasov <ja@ssi.bg>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds a new mode of operation to macvlan, called "source".
It allows one to set a list of allowed mac address, which is used
to match against source mac address from received frames on underlying
interface.
This enables creating mac based VLAN associations, instead of standard
port or tag based. The feature is useful to deploy 802.1x mac based
behavior, where drivers of underlying interfaces doesn't allows that.
Configuration is done through the netlink interface using e.g.:
ip link add link eth0 name macvlan0 type macvlan mode source
ip link add link eth0 name macvlan1 type macvlan mode source
ip link set link dev macvlan0 type macvlan macaddr add 00:11:11:11:11:11
ip link set link dev macvlan0 type macvlan macaddr add 00:22:22:22:22:22
ip link set link dev macvlan0 type macvlan macaddr add 00:33:33:33:33:33
ip link set link dev macvlan1 type macvlan macaddr add 00:33:33:33:33:33
ip link set link dev macvlan1 type macvlan macaddr add 00:44:44:44:44:44
This allows clients with MAC addresses 00:11:11:11:11:11,
00:22:22:22:22:22 to be part of only VLAN associated with macvlan0
interface. Clients with MAC addresses 00:44:44:44:44:44 with only VLAN
associated with macvlan1 interface. And client with MAC address
00:33:33:33:33:33 to be associated with both VLANs.
Based on work of Stefan Gula <steweg@gmail.com>
v8: last version of Stefan Gula for Kernel 3.2.1
v9: rework onto linux-next 2014-03-12 by Michael Braun
add MACADDR_SET command, enable to configure mac for source mode
while creating interface
v10:
- reduce indention level
- rename source_list to source_entry
- use aligned 64bit ether address
- use hash_64 instead of addr[5]
v11:
- rebase for 3.14 / linux-next 20.04.2014
v12
- rebase for linux-next 2014-09-25
Signed-off-by: Michael Braun <michael-dev@fami-braun.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit 412ca1550c ("macvlan: Move broadcasts into a work queue"), the
driver uses tx_queue_len of the master device as the limit of packets enqueuing.
Problem is that virtual drivers have this value set to 0, thus all broadcast
packets were rejected.
Because tx_queue_len was arbitrarily chosen, I replace it with a static limit
of 1000 (also arbitrarily chosen).
CC: Herbert Xu <herbert@gondor.apana.org.au>
Reported-by: Thibaut Collet <thibaut.collet@6wind.com>
Suggested-by: Thibaut Collet <thibaut.collet@6wind.com>
Tested-by: Thibaut Collet <thibaut.collet@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
I cannot move a macvlan interface created on top of a bonding interface
to a different namespace:
% ip netns add dummy0
% ip link add link bond0 mac0 type macvlan
% ip link set mac0 netns dummy0
RTNETLINK answers: Invalid argument
%
The problem seems to be that commit f939981492 ("bonding: Don't allow
bond devices to change network namespaces.") sets NETIF_F_NETNS_LOCAL
on bonding interfaces, and commit 797f87f83b ("macvlan: fix netdev
feature propagation from lower device") causes macvlan interfaces
to inherit its features from the lower device.
NETIF_F_NETNS_LOCAL should not be inherited from the lower device
by a macvlan.
Patch tested on 3.16.
Signed-off-by: Francesco Ruggeri <fruggeri@arista.com>
Acked-by: Cong Wang <cwang@twopensource.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, macvlan code restricts multicast and unicast
filter setting only to passthru devices. As a result,
if a guest using macvtap wants to receive multicast
traffic, it has to set IFF_ALLMULTI or IFF_PROMISC.
This patch makes it possible to use the fdb interface
to add multicast addresses to the filter thus allowing
a guest to receive only targeted multicast traffic.
CC: John Fastabend <john.r.fastabend@intel.com>
CC: Michael S. Tsirkin <mst@redhat.com>
CC: Jason Wang <jasowang@redhat.com>
Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Macvlan devices do not initialize vlan_features. As a result,
any vlan devices configured on top of macvlans perform very poorly.
Initialize vlan_features based on the vlan features of the lower-level
device.
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
net/core/rtnetlink.c
net/core/skbuff.c
Both conflicts were very simple overlapping changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
unregister_netdevice_many() API is error prone and we had too
many bugs because of dangling LIST_HEAD on stacks.
See commit f87e6f4793 ("net: dont leave active on stack LIST_HEAD")
In fact, instead of making sure no caller leaves an active list_head,
just force a list_del() in the callee. No one seems to need to access
the list after unregister_netdevice_many()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Bonding and team drivers generate specific events during failover
that trigger switch updates. When a macvlan device is configured
on top of bonding, we want switches to learn about the macvlan
devices as well. This patch adds a handler to macvlan driver to
propagate these events to all macvlan devices. We let the generic
inetdev event handler do the work.
This allows macvlan to operated correctly over active-backup
mode bond.
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add netpoll support to macvlan devices. Based on the netpoll support in the 802.1q vlan code.
Tested and macvlan could work well with netconsole.
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The macvlan dev should always have the same mac address like lowerdev
when in the passthru mode, change the mac address alone will break the
work mechanism, so when the lowerdev or macvlan mac address changes,
we should propagate the changes to another dev.
v1->v2: Allow macvlan dev to change mac address for passthru mode and propagate to
lowerdev.
v2->v3: Don't set the mac address to the lower dev's unicast address for
passthru mode when mac address changes.
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/bonding/bond_alb.c
drivers/net/ethernet/altera/altera_msgdma.c
drivers/net/ethernet/altera/altera_sgdma.c
net/ipv6/xfrm6_output.c
Several cases of overlapping changes.
The xfrm6_output.c has a bug fix which overlaps the renaming
of skb->local_df to skb->ignore_df.
In the Altera TSE driver cases, the register access cleanups
in net-next overlapped with bug fixes done in net.
Similarly a bug fix to send ALB packets in the bonding driver using
the right source address overlaps with cleanups in net-next.
Signed-off-by: David S. Miller <davem@davemloft.net>
The port->count was used to count the number of macvlan devs
in the same port, but the list vlans could play the same role
to do that, so free the port if the list vlans is empty and
no need to use the parameter count.
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When the physical MTU changes we should ensure that all existing MACVLAN
dev MTU do not exceed the new lowerdev MTU. This patch adds that
propagation.
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Reviewed-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>
Clearing the IFF_ALLMULTI flag on a down interface could cause an allmulti
overflow on the underlying interface.
Attempting the set IFF_ALLMULTI on the underlying interface would cause an
error and the log message:
"allmulti touches root, set allmulti failed."
Signed-off-by: Peter Christensen <pch@ordbogen.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/ethernet/altera/altera_sgdma.c
net/netlink/af_netlink.c
net/sched/cls_api.c
net/sched/sch_api.c
The netlink conflict dealt with moving to netlink_capable() and
netlink_ns_capable() in the 'net' tree vs. supporting 'tc' operations
in non-init namespaces. These were simple transformations from
netlink_capable to netlink_ns_capable.
The Altera driver conflict was simply code removal overlapping some
void pointer cast cleanups in net-next.
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit 12a2856b60.
The commit above doesn't appear to be necessary any more as the
checksums appear to be correctly computed/validated.
Additionally the above commit breaks kvm configurations where
one VM is using a device that support checksum offload (virtio) and
the other VM does not.
In this case, packets leaving virtio device will have CHECKSUM_PARTIAL
set. The packets is forwarded to a macvtap that has offload features
turned off. Since we use CHECKSUM_UNNECESSARY, the host does does not
update the checksum and thus a bad checksum is passed up to
the guest.
CC: Daniel Lezcano <daniel.lezcano@free.fr>
CC: Patrick McHardy <kaber@trash.net>
CC: Andrian Nord <nightnord@gmail.com>
CC: Eric Dumazet <eric.dumazet@gmail.com>
CC: Michael S. Tsirkin <mst@redhat.com>
CC: Jason Wang <jasowang@redhat.com>
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The recent patch that moved broadcasts to process context added
a couple of bugs on the error path where we may dereference NULL
or leak an skb. This patch fixes them.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently broadcasts are handled in network RX context, where
the packets are sent through netif_rx. This means that the number
of macvlans will be constrained by the capacity of netif_rx.
For example, setting up 4096 macvlans practically causes all
broadcast packets to be dropped as the default netif_rx queue
size simply can't handle 4096 skbs being stuffed into it all
at once.
Fundamentally, we need to ensure that the amount of work handled
in each netif_rx backlog run is constrained. As broadcasts are
anything but constrained, it either needs to be limited per run
or moved to process context.
This patch picks the second option and moves all broadcast handling
bar the trivial case of packets going to a single interface into
a work queue. Obviously there also needs to be a limit on how
many broadcast packets we postpone in this way. I've arbitrarily
chosen tx_queue_len of the master device as the limit (act_mirred
also happens to use this parameter in a similar way).
In order to ensure we don't exceed the backlog queue we will use
netif_rx_ni instead of netif_rx for broadcast packets.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Thanks,
Signed-off-by: David S. Miller <davem@davemloft.net>
Replace the bh safe variant with the hard irq safe variant.
We need a hard irq safe variant to deal with netpoll transmitting
packets from hard irq context, and we need it in most if not all of
the places using the bh safe variant.
Except on 32bit uni-processor the code is exactly the same so don't
bother with a bh variant, just have a hard irq safe variant that
everyone can use.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/wireless/ath/ath9k/recv.c
drivers/net/wireless/mwifiex/pcie.c
net/ipv6/sit.c
The SIT driver conflict consists of a bug fix being done by hand
in 'net' (missing u64_stats_init()) whilst in 'net-next' a helper
was created (netdev_alloc_pcpu_stats()) which takes care of this.
The two wireless conflicts were overlapping changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
Macvlan currently inherits all of its features from the lower
device. When lower device disables offload support, this causes
macvlan to disable offload support as well. This causes
performance regression when using macvlan/macvtap in bridge
mode.
It can be easily demonstrated by creating 2 namespaces using
macvlan in bridge mode and running netperf between them:
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.0.0.1 () port 0 AF_INET
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 16384 16384 20.00 1204.61
To restore the performance, we add software offload features
to the list of "always_on" features for macvlan. This way
when a namespace or a guest using macvtap initially sends a
packet, this packet will not be segmented at macvlan level.
It will only be segmented when macvlan sends the packet
to the lower device.
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.0.0.1 () port 0 AF_INET
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 16384 16384 20.00 5507.35
Fixes: 6acf54f1cf (macvtap: Add support of packet capture on macvtap device.)
Fixes: 797f87f83b (macvlan: fix netdev feature propagation from lower device)
CC: Florian Westphal <fw@strlen.de>
CC: Christian Borntraeger <borntraeger@de.ibm.com>
CC: Jason Wang <jasowang@redhat.com>
CC: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/bonding/bond_3ad.h
drivers/net/bonding/bond_main.c
Two minor conflicts in bonding, both of which were overlapping
changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
There are many drivers calling alloc_percpu() to allocate pcpu stats
and then initializing ->syncp. So just introduce a helper function for them.
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
rtnl_newlink() doesn't unregister it for us on failure.
Cc: Patrick McHardy <kaber@trash.net>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Cong Wang <cwang@twopensource.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, the tx queue were selected implicitly in ndo_dfwd_start_xmit(). The
will cause several issues:
- NETIF_F_LLTX were removed for macvlan, so txq lock were done for macvlan
instead of lower device which misses the necessary txq synchronization for
lower device such as txq stopping or frozen required by dev watchdog or
control path.
- dev_hard_start_xmit() was called with NULL txq which bypasses the net device
watchdog.
- dev_hard_start_xmit() does not check txq everywhere which will lead a crash
when tso is disabled for lower device.
Fix this by explicitly introducing a new param for .ndo_select_queue() for just
selecting queues in the case of l2 forwarding offload. netdev_pick_tx() was also
extended to accept this parameter and dev_queue_xmit_accel() was used to do l2
forwarding transmission.
With this fixes, NETIF_F_LLTX could be preserved for macvlan and there's no need
to check txq against NULL in dev_hard_start_xmit(). Also there's no need to keep
a dedicated ndo_dfwd_start_xmit() and we can just reuse the code of
dev_queue_xmit() to do the transmission.
In the future, it was also required for macvtap l2 forwarding support since it
provides a necessary synchronization method.
Cc: John Fastabend <john.r.fastabend@intel.com>
Cc: Neil Horman <nhorman@tuxdriver.com>
Cc: e1000-devel@lists.sourceforge.net
Signed-off-by: Jason Wang <jasowang@redhat.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
L2 fowarding offload will bypass the rx handler of real device. This will make
the packet could not be forwarded to macvtap device. Another problem is the
dev_hard_start_xmit() called for macvtap does not have any synchronization.
Fix this by forbidding L2 forwarding for macvtap.
Cc: John Fastabend <john.r.fastabend@intel.com>
Cc: Neil Horman <nhorman@tuxdriver.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Acked-by: John Fastabend <john.r.fastabend@intel.com.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_pf.c
net/ipv6/ip6_tunnel.c
net/ipv6/ip6_vti.c
ipv6 tunnel statistic bug fixes conflicting with consolidation into
generic sw per-cpu net stats.
qlogic conflict between queue counting bug fix and the addition
of multiple MAC address support.
Signed-off-by: David S. Miller <davem@davemloft.net>
