We introduce the new "End.DT46" action for supporting the SRv6 End.DT46
Behavior in iproute2.
The SRv6 End.DT46 Behavior, defined in RFC 8986 [1] section 4.8, can be
used to implement L3 VPNs based on Segment Routing over IPv6 networks in
multi-tenants environments and it is capable of handling both IPv4 and
IPv6 tenant traffic at the same time.
The SRv6 End.DT46 Behavior decapsulates the received packets and it
performs the IPv4 or IPv6 routing lookup in the routing table of the
tenant.
As for the End.DT4 and for the End.DT6 in VRF mode, the SRv6 End.DT46
Behavior leverages a VRF device in order to force the routing lookup into
the associated routing table using the "vrftable" attribute.
To make the End.DT46 work properly, it must be guaranteed that the
routing table used for routing lookup operations is bound to one and
only one VRF during the tunnel creation. Such constraint has to be
enforced by enabling the VRF strict_mode sysctl parameter, i.e.:
$ sysctl -wq net.vrf.strict_mode=1
Note that the same approach is used for the End.DT4 Behavior and for the
End.DT6 Behavior in VRF mode.
An SRv6 End.DT46 Behavior instance can be created as follows:
$ ip -6 route add 2001:db8::1 encap seg6local action End.DT46 vrftable 100 dev vrf100
Standard Output:
$ ip -6 route show 2001:db8::1
2001:db8::1 encap seg6local action End.DT46 vrftable 100 dev vrf100 metric 1024 pref medium
JSON Output:
$ ip -6 -j -p route show 2001:db8::1
[ {
"dst": "2001:db8::1",
"encap": "seg6local",
"action": "End.DT46",
"vrftable": 100,
"dev": "vrf100",
"metric": 1024,
"flags": [ ],
"pref": "medium"
} ]
This patch updates the route.8 man page and the ip route help with the
information related to End.DT46.
Considering that the same information was missing for the SRv6 End.DT4 and
the End.DT6 Behaviors, we have also added it.
[1] https://www.rfc-editor.org/rfc/rfc8986.html#name-enddt46-decapsulation-and-s
Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>
Signed-off-by: Paolo Lungaroni <paolo.lungaroni@uniroma2.it>
Signed-off-by: David Ahern <dsahern@kernel.org>
Dmytro Linkin says:
====================
Series implements devlink rate commands, which are:
- Dump particular or all rate objects (JSON or non-JSON)
- Add/Delete node rate object
- Set tx rate share/max values for rate object
- Set/Unset parent rate object for other rate object
Examples:
Display all rate objects:
# devlink port function rate show
pci/0000:03:00.0/1 type leaf parent some_group
pci/0000:03:00.0/2 type leaf tx_share 12Mbit
pci/0000:03:00.0/some_group type node tx_share 1Gbps tx_max 5Gbps
Display leaf rate object bound to the 1st devlink port of the
pci/0000:03:00.0 device:
# devlink port function rate show pci/0000:03:00.0/1
pci/0000:03:00.0/1 type leaf
Display node rate object with name some_group of the pci/0000:03:00.0
device:
# devlink port function rate show pci/0000:03:00.0/some_group
pci/0000:03:00.0/some_group type node
Display leaf rate object rate values using IEC units:
# devlink -i port function rate show pci/0000:03:00.0/2
pci/0000:03:00.0/2 type leaf 11718Kibit
Display pci/0000:03:00.0/2 leaf rate object as pretty JSON output:
# devlink -jp port function rate show pci/0000:03:00.0/2
{
"rate": {
"pci/0000:03:00.0/2": {
"type": "leaf",
"tx_share": 1500000
}
}
}
Create node rate object with name "1st_group" on pci/0000:03:00.0 device:
# devlink port function rate add pci/0000:03:00.0/1st_group
Create node rate object with specified parameters:
# devlink port function rate add pci/0000:03:00.0/2nd_group \
tx_share 10Mbit tx_max 30Mbit parent 1st_group
Set parameters to the specified leaf rate object:
# devlink port function rate set pci/0000:03:00.0/1 \
tx_share 2Mbit tx_max 10Mbit
Set leaf's parent to "1st_group":
# devlink port function rate set pci/0000:03:00.0/1 parent 1st_group
Unset leaf's parent:
# devlink port function rate set pci/0000:03:00.0/1 noparent
Delete node rate object:
# devlink port function rate del pci/0000:03:00.0/2nd_group
Rate values can be specified in bits or bytes per second (bit|bps), with
any SI (k, m, g, t) or IEC (ki, mi, gi, ti) prefix. Bare number means
bits per second. Units also printed in "show" command output, but not
necessarily the same which were specified with "set" or "add" command.
-i/--iec switch force output in IEC units. JSON output always print
values as bytes per sec.
====================
Signed-off-by: David Ahern <dsahern@kernel.org>
Implement user commands to manage devlink port func rate objects.
List all rate commands:
$ devlink port func rate help
or just
$ devlink port func rate
To list all OR particular rate object:
$ devlink port func rate show
pci/0000:03:00.0/some_group: type node
pci/0000:03:00.0/0: type leaf
pci/0000:03:00.0/1: type leaf
$ devlink prot func rate show pci/0000:03:00.0/1
pci/0000:03:00.0/0: type leaf
$ devlink prot func rate show pci/0000:03:00.0/some_group
pci/0000:03:00.0/some_group: type node
Rate object of type "leaf" created by it's driver where name is the name
of corresponding devlink port. Rate object of type "node" represents
rate group created by the user using commands:
$ devlink port func rate add pci/0000:03:00.0/some_group
or with defining tx rate limits
$ devlink port func rate add pci/0000:03:00.0/some_group \
tx_shara 10kbit tx_max 100mbit
NOTE: node name cannot be a decimal value because it conflicts with
devlink port indexes.
To delete node object:
$ devlink port func rate del pci/0000:03:00.0/some_group
Set rate limits of existing rate object:
$ devlink prot func rate set pci/0000:03:00.0/0 \
tx_share 5MBps tx_max 25GBps
$ devlink prot func rate set pci/0000:03:00.0/some_group \
tx_share 0
Both SET and ADD commands accept any units of rates defined in IEC
60027-2 standard.
NOTE: rate value 0 means that rate is unlimited. Such value is also
ommited in show command output.
NOTE: In SHOW command output rate values will be printed with suffixes
as well, but in JSON output they are always units of Bps.
Set or unset parent of existing rate object:
$ devlink prot func rate set pci/0000:03:00.0/0 parent some_group
$ devlink port func rate set pci/0000:03:00.0/0 noparent
NOTE: Setting parent to empty ("") name due to kernel logic means unset
parent and shouldn't be used to avoid unexpected parent unsets.
Signed-off-by: Dmytro Linkin <dlinkin@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Every handler argument validated in two steps, first of which, form
checking, expects identifier is few words separated by slashes.
For device and region handlers just checked if identifier have expected
number of slashes.
Add generic function to do that and make code cleaner & consistent.
Signed-off-by: Dmytro Linkin <dlinkin@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
A user optionally provides the external controller number when user
wants to create devlink port for the external controller.
An example on eswitch system:
$ devlink dev eswitch set pci/0033:01:00.0 mode switchdev
$ devlink port show
pci/0033:01:00.0/196607: type eth netdev enP51p1s0f0np0 flavour physical port 0 splittable false
pci/0033:01:00.0/131072: type eth netdev eth0 flavour pcipf controller 1 pfnum 0 external true splittable false
function:
hw_addr 00:00:00:00:00:00
$ devlink port add pci/0033:01:00.0 flavour pcisf pfnum 0 sfnum 77 controller 1
pci/0033:01:00.0/163840: type eth netdev eth1 flavour pcisf controller 1 pfnum 0 sfnum 77 external true splittable false
function:
hw_addr 00:00:00:00:00:00 state inactive opstate detached
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
There are more and more global environment variables that land everywhere
in configure, which is making user hard to know which one does what.
Using command-line options would make it easier for users to learn or
remember the config options.
This patch converts the INCLUDE variable to command option first. Check
if the first variable has '-' to compile with the old INCLUDE path
setting method.
Signed-off-by: Hangbin Liu <haliu@redhat.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Add support for matching on ct_state flag related.
The related state indicates a packet is associated with an existing
connection.
Example:
$ tc filter add dev ens1f0_0 ingress prio 1 chain 1 proto ip flower \
ct_state -est-rel+trk \
action mirred egress redirect dev ens1f0_1
$ tc filter add dev ens1f0_0 ingress prio 1 chain 1 proto ip flower \
ct_state +rel+trk \
action mirred egress redirect dev ens1f0_1
Signed-off-by: Ariel Levkovich <lariel@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
With commit d5e6ee0dac the usage of functions name_to_handle_at() and
open_by_handle_at() are introduced. But these function are not available
e.g. in uclibc-ng < 1.0.35. To have a backward compatibility check for the
availability in the configure script and in case of absence do a direct
syscall.
Fixes: d5e6ee0dac ("ss: introduce cgroup2 cache and helper functions")
Cc: Dmitry Yakunin <zeil@yandex-team.ru>
Cc: Petr Vorel <petr.vorel@gmail.com>
Signed-off-by: Heiko Thiery <heiko.thiery@gmail.com>
Reviewed-by: Petr Vorel <petr.vorel@gmail.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
config.mk needs to be re-generated any time configure is changed.
Rename the existing make target and add a check that the config.mk
file needs to exist and must be newer than configure script.
Signed-off-by: David Ahern <dsahern@kernel.org>
Reviewed-by: Petr Vorel <petr.vorel@gmail.com>
Tested-by: Petr Vorel <petr.vorel@gmail.com>
We introduce the "count" optional attribute for supporting counters in SRv6
Behaviors as defined in [1], section 6. For each SRv6 Behavior instance,
counters defined in [1] are:
- the total number of packets that have been correctly processed;
- the total amount of traffic in bytes of all packets that have been
correctly processed;
In addition, we introduce a new counter that counts the number of packets
that have NOT been properly processed (i.e. errors) by an SRv6 Behavior
instance.
Each SRv6 Behavior instance can be configured, at the time of its creation,
to make use of counters specifing the "count" attribute as follows:
$ ip -6 route add 2001:db8::1 encap seg6local action End count dev eth0
per-behavior counters can be shown by adding "-s" to the iproute2 command
line, i.e.:
$ ip -s -6 route show 2001:db8::1
2001:db8::1 encap seg6local action End packets 0 bytes 0 errors 0 dev eth0
[1] https://www.rfc-editor.org/rfc/rfc8986.html#name-counters
v2:
- add help and route.8 man page updates
Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>
Signed-off-by: Paolo Lungaroni <paolo.lungaroni@uniroma2.it>
Signed-off-by: David Ahern <dsahern@kernel.org>
When a wrong value is provided for "burst" or "cburst" parameters, the
resulting error message is unclear and can be misleading:
$ tc class add dev dummy0 parent 1: classid 1:1 htb rate 100KBps burst errtrigger
Illegal "buffer"
The message claims an illegal "buffer" is provided, but neither the
inline help nor the man page list "buffer" among the htb parameters, and
the only way to know that "burst", "maxburst" and "buffer" are synonyms
is to look into tc/q_htb.c.
This commit tries to improve this simply changing the error string to
the parameter name provided in the user-given command, clearly pointing
out where the wrong value is.
$ tc class add dev dummy0 parent 1: classid 1:1 htb rate 100KBps burst errtrigger
Illegal "burst"
$ tc class add dev dummy0 parent 1: classid 1:1 htb rate 100Kbps maxburst errtrigger
Illegal "maxburst"
Reported-by: Sebastian Mitterle <smitterl@redhat.com>
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
tipc segfaults when called with an abnormally long key:
$ tipc node set key 0123456789abcdef0123456789abcdef0123456789abcdef
*** buffer overflow detected ***: terminated
Fix this returning an error if key length is longer than
TIPC_AEAD_KEYLEN_MAX.
Fixes: 24bee3bf97 ("tipc: add new commands to set TIPC AEAD key")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
tipc segfaults when called with an abnormally long algname:
$ tipc node set key 0x1234 algname supercalifragilistichespiralidososupercalifragilistichespiralidoso
*** buffer overflow detected ***: terminated
Fix this returning an error if provided algname is longer than
TIPC_AEAD_ALG_NAME.
Fixes: 24bee3bf97 ("tipc: add new commands to set TIPC AEAD key")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
When receiving a result from first query to netlink, we may exec
a another query inside the callback. If calling this sub-routine
in the same socket, it will be discarded the result from previous
exection.
To avoid this we perform a nested query in separate socket.
Fixes: 2021028306 ("tipc: use the libmnl functions in lib/mnl_utils.c")
Signed-off-by: Hoang Le <hoang.h.le@dektech.com.au>
Acked-by: Jon Maloy <jmaloy@redhat.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Linux kernel commit b8392808eb3fc28e ("sch_cake: add RFC 8622 LE PHB
support to CAKE diffserv handling") added packets with LE diffserv to
the Bulk priority tin. Update the documentation to reflect this change.
Signed-off-by: Tyson Moore <tyson@tyson.me>
Signed-off-by: David Ahern <dsahern@kernel.org>
main() dinamically allocates dcb, but when dcb_help() is called it
returns without freeing it.
Fix this using a goto, as it is already done in the same function.
Fixes: 67033d1c1c ("Add skeleton of a new tool, dcb")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Reviewed-by: Petr Machata <me@pmachata.org>
Signed-off-by: David Ahern <dsahern@kernel.org>
dcb_cmd_app_show() is supposed to return EINVAL if an incorrect argument
is provided.
Fixes: 8e9bed1493 ("dcb: Add a subtool for the DCB APP object")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Reviewed-by: Petr Machata <me@pmachata.org>
Signed-off-by: David Ahern <dsahern@kernel.org>
In function bpf_obj_open, if bpf_fetch_prog_arg() return an error, we
end up in the out: path with a negative value for fd, and pass it to
close.
Avoid this checking for fd to be positive.
Fixes: 32e93fb7f6 ("{f,m}_bpf: allow for sharing maps")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Checking for nbands to be at least 1 at this point is useless. Indeed:
- ets requires "bands", "quanta" or "strict" to be specified
- if "bands" is specified, nbands cannot be negative, see parse_nbands()
- if "strict" is specified, nstrict cannot be negative, see
parse_nbands()
- if "quantum" is specified, nquanta cannot be negative, see
parse_quantum()
- if "bands" is not specified, nbands is set to nstrict+nquanta
- the previous if statement takes care of the case when none of them are
specified and nbands is 0, terminating execution.
Thus nbands cannot be < 1 at this point and this code cannot be executed.
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Gal Pressman says:
====================
This is the userspace part for the new copy-on-fork attribute added to
the get sys netlink command.
The new attribute indicates that the kernel copies DMA pages on fork,
hence fork support through madvise and MADV_DONTFORK is not needed.
Kernel series was merged:
https://lore.kernel.org/linux-rdma/20210418121025.66849-1-galpress@amazon.com/
====================
Signed-off-by: David Ahern <dsahern@kernel.org>
The new attribute indicates that the kernel copies DMA pages on fork,
hence fork support through madvise and MADV_DONTFORK is not needed.
If the attribute is not reported (expected on older kernels),
copy-on-fork is disabled.
Example:
$ rdma sys
netns shared copy-on-fork on
Signed-off-by: Gal Pressman <galpress@amazon.com>
Acked-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
When add address with port, it is mean to send an ADD_ADDR to remote,
so it must have flag signal set.
Fixes: 42fbca91cd ("mptcp: add support for port based endpoint")
Signed-off-by: Jianguo Wu <wujianguo@chinatelecom.cn>
Acked-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David Ahern <dsahern@kernel.org>
The default behavior for source MACVLAN is to duplicate packets to
appropriate type source devices, and then do the normal destination MACVLAN
flow. This patch adds an option to skip destination MACVLAN processing if
any matching source MACVLAN device has the option set.
This allows setting up a "catch all" device for source MACVLAN: create one
or more devices with type source nodst, and one device with e.g. type vepa,
and incoming traffic will be received on exactly one device.
Signed-off-by: Jethro Beekman <kernel@jbeekman.nl>
Signed-off-by: David Ahern <dsahern@kernel.org>
Leon Romanovsky says:
====================
This is the user space part of already accepted to the kernel series
that extends RDMA netlink interface to return uverbs context and SRQ
information.
The accepted kernel series can be seen here:
https://lore.kernel.org/linux-rdma/20210422133459.GA2390260@nvidia.com/
====================
Signed-off-by: David Ahern <dsahern@kernel.org>
Sample output:
$ rdma res show srq
dev ibp8s0f0 srqn 0 type BASIC pdn 3 comm [ib_ipoib]
dev ibp8s0f0 srqn 4 type BASIC lqpn 125-128,130-140 pdn 9 pid 3581 comm ibv_srq_pingpon
dev ibp8s0f0 srqn 5 type BASIC lqpn 141-156 pdn 10 pid 3584 comm ibv_srq_pingpon
dev ibp8s0f0 srqn 6 type BASIC lqpn 157-172 pdn 11 pid 3590 comm ibv_srq_pingpon
dev ibp8s0f1 srqn 0 type BASIC pdn 3 comm [ib_ipoib]
dev ibp8s0f1 srqn 1 type BASIC lqpn 329-344 pdn 4 pid 3586 comm ibv_srq_pingpon
$ rdma res show srq lqpn 126-141
dev ibp8s0f0 srqn 4 type BASIC lqpn 126-128,130-140 pdn 9 pid 3581 comm ibv_srq_pingpon
dev ibp8s0f0 srqn 5 type BASIC lqpn 141 pdn 10 pid 3584 comm ibv_srq_pingpon
$ rdma res show srq lqpn 127
dev ibp8s0f0 srqn 4 type BASIC lqpn 127 pdn 9 pid 3581 comm ibv_srq_pingpon
Reviewed-by: Ido Kalir <idok@nvidia.com>
Reviewed-by: Mark Zhang <markz@mellanox.com>
Signed-off-by: Neta Ostrovsky <netao@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Sample output:
$ rdma res show ctx
dev ibp8s0f0 ctxn 0 pid 980 comm ibv_rc_pingpong
dev ibp8s0f0 ctxn 1 pid 981 comm ibv_rc_pingpong
dev ibp8s0f0 ctxn 2 pid 992 comm ibv_rc_pingpong
dev ibp8s0f1 ctxn 0 pid 984 comm ibv_rc_pingpong
dev ibp8s0f1 ctxn 1 pid 987 comm ibv_rc_pingpong
$ rdma res show ctx dev ibp8s0f1
dev ibp8s0f1 ctxn 0 pid 984 comm ibv_rc_pingpong
dev ibp8s0f1 ctxn 1 pid 987 comm ibv_rc_pingpong
Reviewed-by: Mark Zhang <markz@mellanox.com>
Reviewed-by: Ido Kalir <idok@nvidia.com>
Signed-off-by: Neta Ostrovsky <netao@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Update kernel headers to commit:
99ba0ea616aa ("sfc: adjust efx->xdp_tx_queue_count with the real number of initialized queues")
Signed-off-by: David Ahern <dsahern@kernel.org>
In functions bpf_{send,recv}_map_fds(), when connect fails after a
socket is successfully opened, we return with error missing a close on
the socket.
Fix this closing the socket if opened and using a single return point
for both the functions.
Fixes: 6256f8c9e4 ("tc, bpf: finalize eBPF support for cls and act front-end")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
As stated in the man page(), open returns a non-negative integer as a
file descriptor. Hence, when checking for its return value to be ok, we
should include 0 as a valid value.
This fixes a covscan warning about a missing close() in this function.
Fixes: ecb05c0f99 ("bpf: improve error reporting around tail calls")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
envp_run is dinamically allocated with a malloc, and not freed in the
out: return path. This commit fix it.
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
In functions netns_pids() and netns_identify_pid(), the netns file is
not closed on some error paths.
Fix this using a conditional close and a single return point on both
functions.
Fixes: 44b563269e ("ip-nexthop: support flush by id")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
When I added support for new vlan rtm dumping, I made a mistake in the
output format when there are no vlans on the port. This patch fixes it by
not printing ports without vlan entries (similar to current situation).
Example (no vlans):
$ bridge -d vlan show
port vlan-id
Fixes: e5f87c8341 ("bridge: vlan: add support for the new rtm dump call")
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
The 'ip' utility hardcodes the assumption of being a 2-char command, where
any follow-on characters are passed as an argument:
$ ./ip-full help
Object "-full" is unknown, try "ip help".
This confusing behaviour isn't seen with 'tc' for example, and was added in
a 2005 commit without documentation. It was noticed during testing of 'ip'
variants built/packaged with different feature sets (e.g. w/o BPF support).
Mitigate the problem by redoing the command without the 2-char assumption
if the follow-on characters fail to parse as a valid command.
Fixes: 351efcde4e ("Update header files to 2.6.14")
Signed-off-by: Tony Ambardar <Tony.Ambardar@gmail.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Paolo Lungaroni
David Ahern
Guillaume Nault
Dmytro Linkin
Parav Pandit
Roi Dayan
Eric Dumazet
Hangbin Liu
Ariel Levkovich
Heiko Thiery
Jakub Kicinski
Andrea Claudi
Hoang Le
Tyson Moore
Gal Pressman
Jianguo Wu
Jethro Beekman
Neta Ostrovsky
Stephen Hemminger
Nikolay Aleksandrov
Tony Ambardar