Ensure the inner IP header is part of the skb's linear data before
setting old_iph. Otherwise, on a non-linear skb, old_iph could point
outside of the packet data.
Unlike classical VXLAN, which always encapsulates Ethernet packets,
VXLAN-GPE can transport IP packets directly. In that case, we need to
look at skb->protocol to figure out if an Ethernet header is present.
Fixes: d342894c5d ("vxlan: virtual extensible lan")
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Link: https://patch.msgid.link/2aa75f6fa62ac9dbe4f16ad5ba75dd04a51d4b99.1718804000.git.gnault@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit f58f45c1e5 ("vxlan: drop packets from invalid src-address")
has recently been added to vxlan mainly in the context of source
address snooping/learning so that when it is enabled, an entry in the
FDB is not being created for an invalid address for the corresponding
tunnel endpoint.
Before commit f58f45c1e5 vxlan was similarly behaving as geneve in
that it passed through whichever macs were set in the L2 header. It
turns out that this change in behavior breaks setups, for example,
Cilium with netkit in L3 mode for Pods as well as tunnel mode has been
passing before the change in f58f45c1e5 for both vxlan and geneve.
After mentioned change it is only passing for geneve as in case of
vxlan packets are dropped due to vxlan_set_mac() returning false as
source and destination macs are zero which for E/W traffic via tunnel
is totally fine.
Fix it by only opting into the is_valid_ether_addr() check in
vxlan_set_mac() when in fact source address snooping/learning is
actually enabled in vxlan. This is done by moving the check into
vxlan_snoop(). With this change, the Cilium connectivity test suite
passes again for both tunnel flavors.
Fixes: f58f45c1e5 ("vxlan: drop packets from invalid src-address")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: David Bauer <mail@david-bauer.net>
Cc: Ido Schimmel <idosch@nvidia.com>
Cc: Nikolay Aleksandrov <razor@blackwall.org>
Cc: Martin KaFai Lau <martin.lau@kernel.org>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: David Bauer <mail@david-bauer.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Simon reported that ndo_change_mtu() methods were never
updated to use WRITE_ONCE(dev->mtu, new_mtu) as hinted
in commit 501a90c945 ("inet: protect against too small
mtu values.")
We read dev->mtu without holding RTNL in many places,
with READ_ONCE() annotations.
It is time to take care of ndo_change_mtu() methods
to use corresponding WRITE_ONCE()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Simon Horman <horms@kernel.org>
Closes: https://lore.kernel.org/netdev/20240505144608.GB67882@kernel.org/
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Sabrina Dubroca <sd@queasysnail.net>
Reviewed-by: Simon Horman <horms@kernel.org>
Acked-by: Shannon Nelson <shannon.nelson@amd.com>
Link: https://lore.kernel.org/r/20240506102812.3025432-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We want to be able to run rtnl_fill_ifinfo() under RCU protection
instead of RTNL in the future.
All rtnl_link_ops->get_link_net() methods already using dev_net()
are ready. I added READ_ONCE() annotations on others.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Cross-merge networking fixes after downstream PR.
Conflicts:
include/linux/filter.h
kernel/bpf/core.c
66e13b615a ("bpf: verifier: prevent userspace memory access")
d503a04f8b ("bpf: Add support for certain atomics in bpf_arena to x86 JIT")
https://lore.kernel.org/all/20240429114939.210328b0@canb.auug.org.au/
No adjacent changes.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ensure the inner IP header is part of skb's linear data before reading
its ECN bits. Otherwise we might read garbage.
One symptom is the system erroneously logging errors like
"vxlan: non-ECT from xxx.xxx.xxx.xxx with TOS=xxxx".
Similar bugs have been fixed in geneve, ip_tunnel and ip6_tunnel (see
commit 1ca1ba465e ("geneve: make sure to pull inner header in
geneve_rx()") for example). So let's reuse the same code structure for
consistency. Maybe we'll can add a common helper in the future.
Fixes: d342894c5d ("vxlan: virtual extensible lan")
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Sabrina Dubroca <sd@queasysnail.net>
Link: https://lore.kernel.org/r/1239c8db54efec341dd6455c77e0380f58923a3c.1714495737.git.gnault@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
VXLAN stores per-VNI statistics using vxlan_vnifilter_count().
These statistics were not updated when arp_reduce() failed its
pskb_may_pull() call.
Use vxlan_vnifilter_count() to update the VNI counter when that
happens.
Fixes: 4095e0e132 ("drivers: vxlan: vnifilter: per vni stats")
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
VXLAN devices update their stats locklessly. Therefore these counters
should either be stored in per-cpu data structures or the updates
should be done using atomic increments.
Since the net_device_core_stats infrastructure is already used in
vxlan_rcv(), use it for the other rx_dropped and tx_dropped counter
updates. Update the other counters atomically using DEV_STATS_INC().
Fixes: d342894c5d ("vxlan: virtual extensible lan")
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of (struct rt6_info *)dst casts, we can use :
#define dst_rt6_info(_ptr) \
container_of_const(_ptr, struct rt6_info, dst)
Some places needed missing const qualifiers :
ip6_confirm_neigh(), ipv6_anycast_destination(),
ipv6_unicast_destination(), has_gateway()
v2: added missing parts (David Ahern)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The VXLAN driver currently does not check if the inner layer2
source-address is valid.
In case source-address snooping/learning is enabled, a entry in the FDB
for the invalid address is created with the layer3 address of the tunnel
endpoint.
If the frame happens to have a non-unicast address set, all this
non-unicast traffic is subsequently not flooded to the tunnel network
but sent to the learnt host in the FDB. To make matters worse, this FDB
entry does not expire.
Apply the same filtering for packets as it is done for bridges. This not
only drops these invalid packets but avoids them from being learnt into
the FDB.
Fixes: d342894c5d ("vxlan: virtual extensible lan")
Suggested-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David Bauer <mail@david-bauer.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Historically, tunnel flags like TUNNEL_CSUM or TUNNEL_ERSPAN_OPT
have been defined as __be16. Now all of those 16 bits are occupied
and there's no more free space for new flags.
It can't be simply switched to a bigger container with no
adjustments to the values, since it's an explicit Endian storage,
and on LE systems (__be16)0x0001 equals to
(__be64)0x0001000000000000.
We could probably define new 64-bit flags depending on the
Endianness, i.e. (__be64)0x0001 on BE and (__be64)0x00010000... on
LE, but that would introduce an Endianness dependency and spawn a
ton of Sparse warnings. To mitigate them, all of those places which
were adjusted with this change would be touched anyway, so why not
define stuff properly if there's no choice.
Define IP_TUNNEL_*_BIT counterparts as a bit number instead of the
value already coded and a fistful of <16 <-> bitmap> converters and
helpers. The two flags which have a different bit position are
SIT_ISATAP_BIT and VTI_ISVTI_BIT, as they were defined not as
__cpu_to_be16(), but as (__force __be16), i.e. had different
positions on LE and BE. Now they both have strongly defined places.
Change all __be16 fields which were used to store those flags, to
IP_TUNNEL_DECLARE_FLAGS() -> DECLARE_BITMAP(__IP_TUNNEL_FLAG_NUM) ->
unsigned long[1] for now, and replace all TUNNEL_* occurrences to
their bitmap counterparts. Use the converters in the places which talk
to the userspace, hardware (NFP) or other hosts (GRE header). The rest
must explicitly use the new flags only. This must be done at once,
otherwise there will be too many conversions throughout the code in
the intermediate commits.
Finally, disable the old __be16 flags for use in the kernel code
(except for the two 'irregular' flags mentioned above), to prevent
any accidental (mis)use of them. For the userspace, nothing is
changed, only additions were made.
Most noticeable bloat-o-meter difference (.text):
vmlinux: 307/-1 (306)
gre.ko: 62/0 (62)
ip_gre.ko: 941/-217 (724) [*]
ip_tunnel.ko: 390/-900 (-510) [**]
ip_vti.ko: 138/0 (138)
ip6_gre.ko: 534/-18 (516) [*]
ip6_tunnel.ko: 118/-10 (108)
[*] gre_flags_to_tnl_flags() grew, but still is inlined
[**] ip_tunnel_find() got uninlined, hence such decrease
The average code size increase in non-extreme case is 100-200 bytes
per module, mostly due to sizeof(long) > sizeof(__be16), as
%__IP_TUNNEL_FLAG_NUM is less than %BITS_PER_LONG and the compilers
are able to expand the majority of bitmap_*() calls here into direct
operations on scalars.
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 3e2f544dd8 ("net: get stats64 if device if driver is
configured") moved the callback to dev_get_tstats64() to net core, so,
unless the driver is doing some custom stats collection, it does not
need to set .ndo_get_stats64.
Since this driver is now relying in NETDEV_PCPU_STAT_TSTATS, then, it
doesn't need to set the dev_get_tstats64() generic .ndo_get_stats64
function pointer.
Signed-off-by: Breno Leitao <leitao@debian.org>
Reviewed-by: Subbaraya Sundeep <sbhatta@marvell.com>
Link: https://lore.kernel.org/r/20240311112437.3813987-2-leitao@debian.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
With commit 34d21de99c ("net: Move {l,t,d}stats allocation to core and
convert veth & vrf"), stats allocation could be done on net core
instead of in this driver.
With this new approach, the driver doesn't have to bother with error
handling (allocation failure checking, making sure free happens in the
right spot, etc). This is core responsibility now.
Remove the allocation in the vxlan driver and leverage the network
core allocation instead.
Signed-off-by: Breno Leitao <leitao@debian.org>
Reviewed-by: Subbaraya Sundeep <sbhatta@marvell.com>
Link: https://lore.kernel.org/r/20240311112437.3813987-1-leitao@debian.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Since commit aed65af1cc ("drivers: make device_type const"), the driver
core can properly handle constant struct device_type. Move the vxlan_type
variable to be a constant structure as well, placing it into read-only
memory which can not be modified at runtime.
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
exit_batch_rtnl() is called while RTNL is held,
and devices to be unregistered can be queued in the dev_kill_list.
This saves one rtnl_lock()/rtnl_unlock() pair per netns
and one unregister_netdevice_many() call.
v4: (Paolo feedback : https://netdev-3.bots.linux.dev/vmksft-net/results/453141/17-udpgro-fwd-sh/stdout )
- Changed vxlan_destroy_tunnels() to use vxlan_dellink()
instead of unregister_netdevice_queue to propely remove
devices from vn->vxlan_list.
- vxlan_destroy_tunnels() can simply iterate one list (vn->vxlan_list)
to find all devices in the most efficient way.
- Moved sanity checks in a separate vxlan_exit_net() method.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Antoine Tenart <atenart@kernel.org>
Link: https://lore.kernel.org/r/20240206144313.2050392-10-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Implement MDB bulk deletion support in the VXLAN driver, allowing MDB
entries to be deleted in bulk according to provided parameters.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch is basically a followup to commit 4e4b1798cc ("vxlan: Add
missing entries to vxlan_get_size()"). All of the attributes in
vxlan_get_size() appear in the same order that they are filled in
vxlan_fill_info() except for IFLA_VXLAN_PORT_RANGE. For consistency, move
that entry to match its order and add a comment, like for all other
entries.
Signed-off-by: Benjamin Poirier <bpoirier@nvidia.com>
Link: https://lore.kernel.org/r/20231027184410.236671-1-bpoirier@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Implement support for MDB get operation by looking up a matching MDB
entry, allocating the skb according to the entry's size and then filling
in the response.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, netlink notifications are sent for individual remote entries
and not for the entire MDB entry itself.
Subsequent patches are going to add MDB get support which will require
the VXLAN driver to reply with an entire MDB entry.
Therefore, as a preparation, factor out a helper to calculate the size
of an individual remote entry. When determining the size of the reply
this helper will be invoked for each remote entry in the MDB entry.
No functional changes intended.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adjust the function's arguments and rename it to allow it to be reused
by future call sites that only have access to 'struct
vxlan_mdb_entry_key', but not to 'struct vxlan_mdb_config'.
No functional changes intended.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
struct nla_policy is usually constant itself, but unless
we make the ranges inside constant we won't be able to
make range structs const. The ranges are not modified
by the core.
Reviewed-by: Johannes Berg <johannes@sipsolutions.net>
Reviewed-by: David Ahern <dsahern@kernel.org>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Link: https://lore.kernel.org/r/20231025162204.132528-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The route lookup can be done now via generic function
udp_tunnel6_dst_lookup() to replace the custom implementation in
vxlan6_get_route().
This is similar to what already done for IPv4 in commit 6f19b2c136
("vxlan: use generic function for tunnel IPv4 route lookup").
Suggested-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The route lookup can be done now via generic function
udp_tunnel_dst_lookup() to replace the custom implementations in
vxlan_get_route().
Note that this patch only touches IPv4, while IPv6 still uses
vxlan6_get_route(). After IPv6 route lookup gets converted as well,
vxlan_xmit_one() can be simplified by removing local variables that
will be passed via "struct ip_tunnel_key", such as remote_ip,
local_ip, flow_flags, label.
Suggested-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for flush VXLAN FDB entries by destination IP. FDB entry is
stored as {MAC, SRC_VNI} + remote. The destination IP is an attribute of
the remote. For multicast entries, the VXLAN driver stores a linked list
of remotes for a given key.
In user space, each remote is represented as a separate entry, so when
flush is sent with filter of 'destination IP', flush only the match
remotes. In case that there are no additional remotes, destroy the entry.
For example, the following are stored as one entry with several remotes:
$ bridge fdb show dev vx10
00:00:00:00:00:00 dst 192.1.1.3 self permanent
00:00:00:00:00:00 dst 192.1.1.1 self permanent
00:00:00:00:00:00 dst 192.1.1.2 self permanent
00:00:00:00:00:00 dst 192.1.1.1 vni 1000 self permanent
When user flush by destination IP x, only the relevant remotes will be
flushed:
$ bridge fdb flush dev vx10 dst 192.1.1.1
$ bridge fdb show dev vx10
00:00:00:00:00:00 dst 192.1.1.3 self permanent
00:00:00:00:00:00 dst 192.1.1.2 self permanent
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for flush VXLAN FDB entries by destination port. FDB entry
is stored as {MAC, SRC_VNI} + remote. The destination port is an attribute
of the remote. For multicast entries, the VXLAN driver stores a linked list
of remotes for a given key.
In user space, each remote is represented as a separate entry, so when
flush is sent with filter of 'destination port', flush only the match
remotes. In case that there are no additional remotes, destroy the entry.
For example, the following are stored as one entry with several remotes:
$ bridge fdb show dev vx10
00:00:00:00:00:00 dst 192.1.1.1 port 1111 vni 2000 self permanent
00:00:00:00:00:00 dst 192.1.1.1 port 1111 vni 3000 self permanent
00:00:00:00:00:00 dst 192.1.1.1 port 2222 vni 2000 self permanent
00:00:00:00:00:00 dst 192.1.1.1 vni 3000 self permanent
When user flush by port x, only the relevant remotes will be flushed:
$ bridge fdb flush dev vx10 port 1111
$ bridge fdb show dev vx10
00:00:00:00:00:00 dst 192.1.1.1 port 2222 vni 2000 self permanent
00:00:00:00:00:00 dst 192.1.1.1 vni 3000 self permanent
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for flush VXLAN FDB entries by destination VNI. FDB entry is
stored as {MAC, SRC_VNI} + remote. The destination VNI is an attribute
of the remote. For multicast entries, the VXLAN driver stores a linked list
of remotes for a given key.
In user space, each remote is represented as a separate entry, so when
flush is sent with filter of 'destination VNI', flush only the match
remotes. In case that there are no additional remotes, destroy the entry.
For example, the following are stored as one entry with several remotes:
$ bridge fdb show dev vx10
00:00:00:00:00:00 dst 192.1.1.1 vni 3000 self permanent
00:00:00:00:00:00 dst 192.1.1.1 vni 4000 self permanent
00:00:00:00:00:00 dst 192.1.1.1 vni 2000 self permanent
00:00:00:00:00:00 dst 192.1.1.2 vni 2000 self permanent
When user flush by VNI x, only the relevant remotes will be flushed:
$ bridge fdb flush dev vx10 vni 2000
$ bridge fdb show dev vx10
00:00:00:00:00:00 dst 192.1.1.1 vni 3000 self permanent
00:00:00:00:00:00 dst 192.1.1.1 vni 4000 self permanent
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for flush VXLAN FDB entries by nexthop ID.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for flush VXLAN FDB entries by source VNI.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The merge commit 9271686937 ("Merge branch 'br-flush-filtering'")
added support for FDB flushing in bridge driver only, the VXLAN driver does
not support such flushing. Extend VXLAN driver to support FDB flushing.
In this commit, add support for flushing with state and flags, which are
the fields that supported in the bridge driver.
Note that bridge driver supports 'NTF_USE' flag, but there is no point to
support this flag for flushing as it is ignored when flags are stored.
'NTF_STICKY' is not relevant for VXLAN driver.
'NTF_ROUTER' is not supported in bridge driver for flush as it is not
relevant for bridge, add it for VXLAN.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, the function vxlan_flush() does not flush the default FDB entry
(an entry with all_zeros_mac and default VNI), as it is deleted at
vxlan_uninit(). When this function will be used for flushing FDB entries
from user space, it will have to flush also the default entry in case that
other parameters match (e.g., VNI, flags).
Extend 'struct vxlan_fdb_flush_desc' to include an indication whether
the default entry should be flushed or not. The default value (false)
indicates to flush it, adjust all the existing callers to set
'.ignore_default_entry' to true, so the current behavior will not be
changed.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function vxlan_flush() gets a boolean called 'do_all' and in case
that it is false, it does not flush entries with state 'NUD_PERMANENT'
or 'NUD_NOARP'. The following patches will add support for FDB flush
with parameters from user space. Make the function more generic, so it
can be used later.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There are some attributes added by vxlan_fill_info() which are not
accounted for in vxlan_get_size(). Add them.
I didn't find a way to trigger an actual problem from this miscalculation
since there is usually extra space in netlink size calculations like
if_nlmsg_size(); but maybe I just didn't search long enough.
Fixes: 3511494ce2 ("vxlan: Group Policy extension")
Fixes: e1e5314de0 ("vxlan: implement GPE")
Fixes: 0ace2ca89c ("vxlan: Use checksum partial with remote checksum offload")
Fixes: f9c4bb0b24 ("vxlan: vni filtering support on collect metadata device")
Signed-off-by: Benjamin Poirier <bpoirier@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function is not called from an atomic context so use GFP_KERNEL
instead of GFP_ATOMIC. The allocation of the per-CPU stats is already
performed with GFP_KERNEL.
Tested using test_vxlan_vnifiltering.sh with CONFIG_DEBUG_ATOMIC_SLEEP.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Link: https://lore.kernel.org/r/20230821141923.1889776-1-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Use the helper functions dev_sw_netstats_rx_add() and
dev_sw_netstats_tx_add() to update stats, which helps to
provide code readability.
Signed-off-by: Li Zetao <lizetao1@huawei.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
In case rhashtable_lookup_insert_fast() fails inside vxlan_vni_add(), the
allocated percpu vni stats are not freed on the error path.
Introduce vxlan_vni_free() which would work as a nice wrapper to free
vxlan_vni_node resources properly.
Found by Linux Verification Center (linuxtesting.org).
Fixes: 4095e0e132 ("drivers: vxlan: vnifilter: per vni stats")
Suggested-by: Ido Schimmel <idosch@idosch.org>
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
In VXLAN-GPE, there may not be an Ethernet header following the VXLAN
header. But in GRO, the vxlan driver calls eth_gro_receive
unconditionally, which means the following header is incorrectly parsed
as Ethernet.
Introduce GPE specific GRO handling.
For better performance, do not check for GPE during GRO but rather
install a different set of functions at setup time.
Fixes: e1e5314de0 ("vxlan: implement GPE")
Reported-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The vxlan_parse_gpe_hdr function extracts the next protocol value from
the GPE header and marks GPE bits as parsed.
In order to be used in the next patch, split the function into protocol
extraction and bit marking. The bit marking is meaningful only in
vxlan_rcv; move it directly there.
Rename the function to vxlan_parse_gpe_proto to reflect what it now
does. Remove unused arguments skb and vxflags. Move the function earlier
in the file to allow it to be called from more places in the next patch.
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
VXLAN-GPE does not add an extra inner Ethernet header. Take that into
account when calculating header length.
This causes problems in skb_tunnel_check_pmtu, where incorrect PMTU is
cached.
In the collect_md mode (which is the only mode that VXLAN-GPE
supports), there's no magic auto-setting of the tunnel interface MTU.
It can't be, since the destination and thus the underlying interface
may be different for each packet.
So, the administrator is responsible for setting the correct tunnel
interface MTU. Apparently, the administrators are capable enough to
calculate that the maximum MTU for VXLAN-GPE is (their_lower_MTU - 36).
They set the tunnel interface MTU to 1464. If you run a TCP stream over
such interface, it's then segmented according to the MTU 1464, i.e.
producing 1514 bytes frames. Which is okay, this still fits the lower
MTU.
However, skb_tunnel_check_pmtu (called from vxlan_xmit_one) uses 50 as
the header size and thus incorrectly calculates the frame size to be
1528. This leads to ICMP too big message being generated (locally),
PMTU of 1450 to be cached and the TCP stream to be resegmented.
The fix is to use the correct actual header size, especially for
skb_tunnel_check_pmtu calculation.
Fixes: e1e5314de0 ("vxlan: implement GPE")
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
VXLAN FDB entries can point to FDB nexthop objects. Each such object
includes the IP address(es) of remote VTEP(s) via which the target host
is accessible. Example:
# ip nexthop add id 1 via 192.0.2.1 fdb
# ip nexthop add id 2 via 192.0.2.17 fdb
# ip nexthop add id 1000 group 1/2 fdb
# bridge fdb add 00:11:22:33:44:55 dev vx0 self static nhid 1000 src_vni 10020
This is useful for EVPN multihoming where a single host can be connected
to multiple VTEPs. The source VTEP will calculate the flow hash of the
skb and forward it towards the IP address of one of the VTEPs member in
the nexthop group.
There are cases where an external entity (e.g., the bridge driver) can
provide not only the tunnel ID (i.e., VNI) of the skb, but also the ID
of the nexthop object via which the skb should be forwarded.
Therefore, in order to support such cases, when the VXLAN device is in
external / collect metadata mode and the tunnel info attached to the skb
is of bridge type, extract the nexthop ID from the tunnel info. If the
ID is valid (i.e., non-zero), forward the skb via the nexthop object
associated with the ID, as if the skb hit an FDB entry associated with
this ID.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
If a packet needs to be encapsulated towards a local destination IP, the
packet will undergo a "local bypass" and be injected into the Rx path as
if it was received by the target VXLAN device without undergoing
encapsulation. If such a device does not exist, the packet will be
dropped.
There are scenarios where we do not want to perform such a bypass, but
instead want the packet to be encapsulated and locally received by a
user space program for post-processing.
To that end, add a new VXLAN device attribute that controls whether a
"local bypass" is performed or not. Default to performing a bypass to
maintain existing behavior.
Signed-off-by: Vladimir Nikishkin <vladimir@nikishkin.pw>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function vxlan_build_gbp_hdr will be used by other modules to build
gbp option in vxlan header according to gbp flags.
Signed-off-by: Gavin Li <gavinl@nvidia.com>
Reviewed-by: Gavi Teitz <gavi@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Acked-by: Saeed Mahameed <saeedm@nvidia.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Remove unused argument (i.e. u32 vxflags) in vxlan_build_gbp_hdr( ) and
vxlan_build_gpe_hdr( ) function arguments.
Signed-off-by: Gavin Li <gavinl@nvidia.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Now that the VXLAN MDB control and data paths are in place we can expose
the VXLAN MDB functionality to user space.
Set the VXLAN MDB net device operations to the appropriate functions,
thereby allowing the rtnetlink code to reach the VXLAN driver.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Integrate MDB support into the Tx path of the VXLAN driver, allowing it
to selectively forward IP multicast traffic according to the matched MDB
entry.
If MDB entries are configured (i.e., 'VXLAN_F_MDB' is set) and the
packet is an IP multicast packet, perform up to three different lookups
according to the following priority:
1. For an (S, G) entry, using {Source VNI, Source IP, Destination IP}.
2. For a (*, G) entry, using {Source VNI, Destination IP}.
3. For the catchall MDB entry (0.0.0.0 or ::), using the source VNI.
The catchall MDB entry is similar to the catchall FDB entry
(00:00:00:00:00:00) that is currently used to transmit BUM (broadcast,
unknown unicast and multicast) traffic. However, unlike the catchall FDB
entry, this entry is only used to transmit unregistered IP multicast
traffic that is not link-local. Therefore, when configured, the catchall
FDB entry will only transmit BULL (broadcast, unknown unicast,
link-local multicast) traffic.
The catchall MDB entry is useful in deployments where inter-subnet
multicast forwarding is used and not all the VTEPs in a tenant domain
are members in all the broadcast domains. In such deployments it is
advantageous to transmit BULL (broadcast, unknown unicast and link-local
multicast) and unregistered IP multicast traffic on different tunnels.
If the same tunnel was used, a VTEP only interested in IP multicast
traffic would also pull all the BULL traffic and drop it as it is not a
member in the originating broadcast domain [1].
If the packet did not match an MDB entry (or if the packet is not an IP
multicast packet), return it to the Tx path, allowing it to be forwarded
according to the FDB.
If the packet did match an MDB entry, forward it to the associated
remote VTEPs. However, if the entry is a (*, G) entry and the associated
remote is in INCLUDE mode, then skip over it as the source IP is not in
its source list (otherwise the packet would have matched on an (S, G)
entry). Similarly, if the associated remote is marked as BLOCKED (can
only be set on (S, G) entries), then skip over it as well as the remote
is in EXCLUDE mode and the source IP is in its source list.
[1] https://datatracker.ietf.org/doc/html/draft-ietf-bess-evpn-irb-mcast#section-2.6
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add an internal flag to indicate whether MDB entries are configured or
not. Set the flag after installing the first MDB entry and clear it
before deleting the last one.
The flag will be consulted by the data path which will only perform an
MDB lookup if the flag is set, thereby keeping the MDB overhead to a
minimum when the MDB is not used.
Another option would have been to use a static key, but it is global and
not per-device, unlike the current approach.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
