The citied commit broke the CLI output and printed ifindex/ifname
instead of dev/link.
Before:
[leonro@vm ~]$ rdma res show qp
link mlx5_0/lqpn 1 type GSI state RTS sq-psn 0 comm ib_core
[leonro@vm ~]$ rdma res show cq
ifindex 0 ifname rocep0s9 cqn 0 cqe 1023 users 2 poll-ctx WORKQUEUE adaptive-moderation on comm ib_core
After:
[leonro@vm ~]$ rdma res show qp
link mlx5_0/- lqpn 1 type GSI state RTS sq-psn 0 comm [ib_core]
[leonro@vm ~]$ rdma res show cq
dev rocep0s9 cqn 0 cqe 1023 users 2 poll-ctx WORKQUEUE adaptive-moderation on comm [ib_core]
It was missed because rdmatool mostly used in JSON mode.
Fixes: b0a688a542 ("rdma: Rewrite custom JSON and prints logic to use common API")
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Owner of kernel resources is printed in different format than user
resources to easy with the reader by simply looking on the name.
The kernel owner will have "[ ]" around the name.
Before this change:
[leonro@vm ~]$ rdma res show qp
link rocep0s9/1 lqpn 1 type GSI state RTS sq-psn 58 comm ib_core
After this change:
[leonro@vm ~]$ rdma res show qp
link rocep0s9/1 lqpn 1 type GSI state RTS sq-psn 58 comm [ib_core]
Fixes: b0a688a542 ("rdma: Rewrite custom JSON and prints logic to use common API")
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Document the new supported criteria of auto mode. Examples:
$ rdma statistic qp set link mlx5_2/1 auto pid on
$ rdma statistic qp set link mlx5_2/1 auto pid,type on
Signed-off-by: Mark Zhang <markz@mellanox.com>
Reviewed-by: Ido Kalir <idok@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
With this new criteria, QPs have different PIDs will be bound to
different counters in auto mode. This can be used in combination with
other criteria like "type". Examples:
$ rdma statistic qp set link mlx5_2/1 auto pid on
$ rdma statistic qp set link mlx5_2/1 auto type,pid on
$ rdma statistic qp set link mlx5_2/1 auto off
$ rdma statistic qp show link mlx5_0 qp-type UD
Signed-off-by: Mark Zhang <markz@mellanox.com>
Reviewed-by: Ido Kalir <idok@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Replace the iproute2 snapshot with a version string which is
autogenerated as part of the build process using git describe.
This will also allow seeing if the version of the command
is built from the same sources is as upstream.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
This flag allows to create SA where sequence number can cycle in
outbound packets if set.
Signed-off-by: Petr Vaněk <pv@excello.cz>
Signed-off-by: David Ahern <dsahern@kernel.org>
Add a new attribute that indicates the port split ability to devlink port.
Expose the attribute to user space as RO value, for example:
$devlink port show swp1
pci/0000:03:00.0/61: type eth netdev swp1 flavour physical port 1
splittable false lanes 1
Signed-off-by: Danielle Ratson <danieller@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Add a new attribute that indicates the port's number of lanes to devlink port.
Expose the attribute to user space as RO value, for example:
$devlink port show swp1
pci/0000:03:00.0/61: type eth netdev swp1 flavour physical port 1 lanes 1
Signed-off-by: Danielle Ratson <danieller@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
bridge json fdb show is printing an incorrect / non-machine readable
value, when using -j (json output) we are expecting machine readable
data that shouldn't require special handling/parsing.
$ bridge -j fdb show | \
python -c \
'import sys,json;print(json.dumps(json.loads(sys.stdin.read()),indent=4))'
[
{
"master": "br0",
"mac": "56:23:28:4f:4f:e5",
"flags": [],
"ifname": "vx0",
"state": "state=0x80" <<<<<<<<< with the patch: "state": "0x80"
}
]
Fixes: c7c1a1ef51 ("bridge: colorize output and use JSON print library")
Signed-off-by: Julien Fortin <julien@cumulusnetworks.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Add space after format specifier in print_string call. Fixes broken
qdisc tests within tdc testing suite. Per suggestion from Petr Machata,
remove a space and change spacing in tc/q_event.c to complete the fix.
Tested fix in tdc using:
./tdc.py -c qdisc
All qdisc RED tests return ok.
Fixes: d0e450438571("tc: q_red: Add support for qevents "mark" and "early_drop")
Signed-off-by: Briana Oursler <briana.oursler@gmail.com>
Tested-by: Petr Machata <petrm@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
In most of cases a user wants to see only the dynamic mac addresses
in the fdb output. But currently the 'fdb show' displays tons of
various self entries, those only waste the output without any useful
goal.
New option 'dynamic' for 'show' and 'get' commands forces display
only relevant records.
Signed-off-by: Anton Danilov <littlesmilingcloud@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
According to 'ip mptcp help', 'endpoint show' can accept no argument:
ip mptcp endpoint show [ id ID ]
It makes sense to print all endpoints when no filter is used.
So here if the following command is used, all endpoints are printed:
ip mptcp endpoint show
Same as:
ip mptcp endpoint
Fixes: 7e0767cd ("add support for mptcp netlink interface")
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Moshe Shemesh says:
====================
Implement commands for interaction with per-port devlink health
reporters. To do this, adapt devlink-health for usage of port handles
with any existing devlink-health subcommands. Add devlink-port health
subcommand as an alias for devlink-health.
====================
Signed-off-by: David Ahern <dsahern@kernel.org>
Add devlink port health show subcommand which displays information about
specified port reporter or all present port reporters as in the example.
Device and port reporters can be distinguished by a handle being used.
Make other devlink-health subcommands be aliased by devlink port health.
Refactor devlink-health commands for usage of port handles in order to
interact with port reporters.
Change devlink health show output to dump information about both device
and port reporters with correct handles.
Example:
$ devlink health show
pci/0000:00:0b.0:
reporter fw
state healthy error 0 recover 0 auto_dump true
reporter fw_fatal
state healthy error 0 recover 0 grace_period 1200000 auto_recover true auto_dump true
pci/0000:00:0b.0/1:
reporter tx
state healthy error 0 recover 0 grace_period 10000 auto_recover true auto_dump true
reporter rx
state healthy error 0 recover 0 grace_period 10000 auto_recover true auto_dump true
$ devlink health show pci/0000:00:0b.0/1 reporter rx
Which is equivalent to:
$ devlink port health show pci/0000:00:0b.0/1 reporter rx
pci/0000:00:0b.0/1:
reporter rx
state healthy error 0 recover 0 grace_period 10000 auto_recover true auto_dump true
$ devlink port health show pci/0000:00:0b.0/1 reporter rx -j --pretty
{
"health": {
"pci/0000:00:0b.0/1": [ {
"reporter": "rx",
"state": "healthy",
"error": 0,
"recover": 0,
"grace_period": 500,
"auto_recover": true,
"auto_dump": true
} ]
}
}
$ devlink health set pci/0000:00:0b.0/1 reporter rx grace_period 5000
Which is equivalent to:
$ devlink port health set pci/0000:00:0b.0/1 reporter rx grace_period 5000
$ devlink port health show pci/0000:00:0b.0/1 reporter rx
pci/0000:00:0b.0/1:
reporter rx
state healthy error 0 recover 0 grace_period 5000 auto_recover true auto_dump true
Signed-off-by: Vladyslav Tarasiuk <vladyslavt@mellanox.com>
Reviewed-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Add a capability of printing port handles for arrays in non-JSON format
in devlink-health manner.
Signed-off-by: Vladyslav Tarasiuk <vladyslavt@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Test the plain MPLS (unicast and multicast) and IP (v4 and v6) modes.
Also test the multiproto option for MPLS and for IP.
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
To improve the usability better use case-insensitive pattern-matching
in ifstat, nstat and ss tools.
Signed-off-by: Anton Danilov <littlesmilingcloud@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
While looking at the estimator code, I noticed an incorrect interval
number printed in raw for the handles. This patch fixes the formatting.
Before patch:
root@bytecenter.fr:~# tc -r filter add dev eth0 ingress estimator
250ms 999ms matchall action police avrate 12mbit conform-exceed drop
[estimator i=4294967294 e=2]
After patch:
root@bytecenter.fr:~# tc -r filter add dev eth0 ingress estimator
250ms 999ms matchall action police avrate 12mbit conform-exceed drop
[estimator i=-2 e=2]
Signed-off-by: Jamie Gloudon <jamie.gloudon@gmx.fr>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Petr Machata says:
====================
When a list of filters at a given block is requested, tc first validates
that the block exists before doing the filter query. Currently the
validation routine checks ingress and egress blocks. But now that blocks
can be bound to qevents as well, qevent blocks should be looked for as
well:
# ip link add up type dummy
# tc qdisc add dev dummy1 root handle 1: \
red min 30000 max 60000 avpkt 1000 qevent early_drop block 100
# tc filter add block 100 pref 1234 handle 102 matchall action drop
# tc filter show block 100
Cannot find block "100"
This patchset fixes this issue:
# tc filter show block 100
filter protocol all pref 1234 matchall chain 0
filter protocol all pref 1234 matchall chain 0 handle 0x66
not_in_hw
action order 1: gact action drop
random type none pass val 0
index 2 ref 1 bind 1
In patch #1, the helpers and necessary infrastructure is introduced,
including a new qdisc_util callback that implements sniffing out bound
blocks in a given qdisc.
In patch #2, RED implements the new callback.
v3:
- Patch #1:
- Do not pass &ctx->found directly to has_block. Do it through a
helper variable, so that the callee does not overwrite the result
already stored in ctx->found.
v2:
- Patch #1:
- In tc_qdisc_block_exists_cb(), do not initialize 'q'.
- Propagate upwards errors from q->has_block.
====================
Signed-off-by: David Ahern <dsahern@kernel.org>
In order for "tc filter show block X" to find a given block, implement the
has_block callback.
Signed-off-by: Petr Machata <petrm@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
When a list of filters at a given block is requested, tc first validates
that the block exists before doing the filter query. Currently the
validation routine checks ingress and egress blocks. But now that blocks
can be bound to qevents as well, qevent blocks should be looked for as
well.
In order to support that, extend struct qdisc_util with a new callback,
has_block. That should report whether, give the attributes in TCA_OPTIONS,
a blocks with a given number is bound to a qevent. In
tc_qdisc_block_exists_cb(), invoke that callback when set.
Add a helper to the tc_qevent module that walks the list of qevents and
looks for a given block. This is meant to be used by the individual qdiscs.
Signed-off-by: Petr Machata <petrm@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
This implement support for MPTCP sockets type, comprising
extended socket info. Note that we need to add an extended
attribute carrying the actual protocol number to the diag
request.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
The XFRMA_IF_ID attribute is set in policies for them to be
associated with an XFRM interface (4.19+).
Add support for getting/deleting policies with this attribute.
For supporting 'deleteall' the XFRMA_IF_ID attribute needs to be
explicitly copied.
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
In commit aed63ae1ac ("ip xfrm: support setting/printing XFRMA_IF_ID attribute in states/policies")
I added the ability to set/print the xfrm interface ID without updating
the man page.
Fixes: aed63ae1ac ("ip xfrm: support setting/printing XFRMA_IF_ID attribute in states/policies")
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Fixes: 5027f233e3 ("tipc: add link broadcast get")
Signed-off-by: Hoang Huu Le <hoang.h.le@dektech.com.au>
Acked-by: Jon Maloy <jmaloy@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
'bridge fdb get' has json support but the json object is never initialized
before patch:
$ bridge -j fdb get 56:23:28:4f:4f:e5 dev vx0
56:23:28:4f:4f:e5 dev vx0 master br0 permanent
$
after patch:
$ bridge -j fdb get 56:23:28:4f:4f:e5 dev vx0 | \
python -c \
'import sys,json;print(json.dumps(json.loads(sys.stdin.read()),indent=4))'
[
{
"master": "br0",
"mac": "56:23:28:4f:4f:e5",
"flags": [],
"ifname": "vx0",
"state": "permanent"
}
]
$
Signed-off-by: Julien Fortin <julien@cumulusnetworks.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
The configure script checks for ipset v6 availability but doesn't test
for v7, which is backward compatible and used on kernel v5.x systems.
Update the script to test for both ipset versions. Without this change,
the tc ematch function em_ipset will be disabled.
Signed-off-by: Tony Ambardar <Tony.Ambardar@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
utils.h is included two times in ipaddress.c, there is no need for that.
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Follow the precedent of other parts of iproute2 follow the example of:
Standard libc headers
Linux headers
Iproute2 support headers
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
The 'fw_load_policy' devlink parameter supports the 'disk' value
since kernel v5.4, seems like there was some oversight in adding
this to iproute, fixed by this patch.
Signed-off-by: Louis Peens <louis.peens@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
When setting a policer to a trap group, a value of "0" will unbind the
currently bound policer from the group.
The behavior is intentional and tested in kernel selftests, so document
it.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Suggested-by: Alex Kushnarov <alexanderk@mellanox.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Add the new "mpls" keyword that can be used to match MPLS fields in
arbitrary Label Stack Entries.
LSEs are introduced by the "lse" keyword and followed by LSE options:
"depth", "label", "tc", "bos" and "ttl". The depth is manadtory, the
other options are optionals.
For example, the following filter drops MPLS packets having two labels,
where the first label is 21 and has TTL 64 and the second label is 22:
$ tc filter add dev ethX ingress proto mpls_uc flower mpls \
lse depth 1 label 21 ttl 64 \
lse depth 2 label 22 bos 1 \
action drop
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Bareudp devices provide a generic L3 encapsulation for tunnelling
different protocols like MPLS, IP, NSH, etc. inside a UDP tunnel.
This patch is based on original work from Martin Varghese:
https://lore.kernel.org/netdev/1570532361-15163-1-git-send-email-martinvarghesenokia@gmail.com/
Examples:
- ip link add dev bareudp0 type bareudp dstport 6635 ethertype mpls_uc
This creates a bareudp tunnel device which tunnels L3 traffic with
ethertype 0x8847 (unicast MPLS traffic). The destination port of the
UDP header will be set to 6635. The device will listen on UDP port 6635
to receive traffic.
- ip link add dev bareudp0 type bareudp dstport 6635 ethertype ipv4 multiproto
Same as the MPLS example, but for IPv4. The "multiproto" keyword allows
the device to also tunnel IPv6 traffic.
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Before this patch check is happened only in case when we try to find
cgroup at cgroup2 mount point.
v2:
- add Fixes line before Signed-off-by (David Ahern)
Fixes: d5e6ee0dac ("ss: introduce cgroup2 cache and helper functions")
Signed-off-by: Dmitry Yakunin <zeil@yandex-team.ru>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
ip(8) accepts -family ipv6 (-6) option at the toplevel. It is
straightforward to support the existing option for modifying listener
on IPv6 addresses.
Maintain the backward compatibility by leaving ip fou -6 flag
implemented, while it's removed from the usage message.
Signed-off-by: Sorah Fukumori <her@sorah.jp>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Before can be possible show only all qeueue disciplines on an interface.
There wasn't a way to get the qdisc info by handle or parent, only full
dump of the disciplines with a following grep/sed usage.
Now new and old options work as expected to filter a qdisc by handle or
parent.
Full syntax of the qdisc show command:
tc qdisc { show | list } [ dev STRING ] [ QDISC_ID ] [ invisible ]
QDISC_ID := { root | ingress | handle QHANDLE | parent CLASSID }
This change doesn't require any changes in the kernel.
Signed-off-by: Anton Danilov <littlesmilingcloud@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Stephen Hemminger
Leon Romanovsky
Mark Zhang
David Ahern
Vasundhara Volam
Petr Vaněk
Danielle Ratson
Julien Fortin
Briana Oursler
Anton Danilov
Matthieu Baerts
Vladyslav Tarasiuk
Guillaume Nault
Jamie Gloudon
Petr Machata
Paolo Abeni
Eyal Birger
Hoang Huu Le
Tony Ambardar
Andrea Claudi
Louis Peens
Ido Schimmel
Dmitry Yakunin
Sorah Fukumori