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>
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>
Use a single font macro for a single argument.
Remove unnecessary quotes for a single-font macro.
Join two lines into one.
The output of "nroff" and "groff" is unchanged.
Signed-off-by: Bjarni Ingi Gislason <bjarniig@rhi.hi.is>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Use a single-font macro for one argument.
Remove unnecessary quotes for a single font macro.
Join some lines into one.
The output of "nroff" and "groff" is unchanged, except for a font
change in two lines.
Signed-off-by: Bjarni Ingi Gislason <bjarniig@rhi.hi.is>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Use a single-font macro for a single argument
Signed-off-by: Bjarni Ingi Gislason <bjarniig@rhi.hi.is>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Use a single-font macro for a single argument.
Signed-off-by: Bjarni Ingi Gislason <bjarniig@rhi.hi.is>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Typeset section numbers in roman font, see man-pages(7).
###
Details:
Output is from: test-groff -b -mandoc -T utf8 -rF0 -t -w w -z
[ "test-groff" is a developmental version of "groff" ]
<./man/man3/libnetlink.3>:53 (macro BR): only 1 argument, but more are expected
<./man/man3/libnetlink.3>:132 (macro BR): only 1 argument, but more are expected
<./man/man3/libnetlink.3>:134 (macro BR): only 1 argument, but more are expected
<./man/man3/libnetlink.3>:197 (macro BR): only 1 argument, but more are expected
<./man/man3/libnetlink.3>:198 (macro BR): only 1 argument, but more are expected
Signed-off-by: Bjarni Ingi Gislason <bjarniig@rhi.hi.is>
Leon Romanovsky says:
====================
The following series adds support to get the RDMA resource data in RAW
format. The main motivation for doing this is to enable vendors to
return the entire QP/CQ/MR data without a need from the vendor to set
each field separately.
====================
Signed-off-by: David Ahern <dsahern@kernel.org>
Add the required support to print MR data in raw format.
Example:
$rdma res show mr dev mlx5_1 mrn 2 -r -j
[{"ifindex":7,"ifname":"mlx5_1",
"data":[0,4,255,254,0,0,0,0,0,0,0,0,16,28,0,216,...]}]
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Add the required support to print CQ data in raw format.
Example:
$rdma res show cq dev mlx5_2 cqn 1 -r -j
[{"ifindex":8,"ifname":"mlx5_2",
"data":[0,4,255,254,0,0,0,0,0,0,0,0,16,28,...]}]
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Add 'raw' argument to get the resource in raw format.
When RDMA_NLDEV_ATTR_RES_RAW is set in the netlink message,
then the resource fields are in raw format, print it as byte array.
Example:
$rdma res show qp link rocep0s12f0/1 lqpn 1137 -j -r
[{"ifindex":7,"ifname":"mlx5_1","port":1,
"data":[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...]}]
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Update rdma_netlink.h file upto kernel commit 65959522f806
("RDMA: Add support to dump resource tracker in RAW format")
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Petr Machata says:
====================
To allow configuring user-defined actions as a result of inner workings of
a qdisc, a concept of qevents was recently introduced to the kernel.
Qevents are attach points for TC blocks, where filters can be put that are
executed as the packet hits well-defined points in the qdisc algorithms.
The attached blocks can be shared, in a manner similar to clsact ingress
and egress blocks, arbitrary classifiers with arbitrary actions can be put
on them, etc.
For example:
# tc qdisc add dev eth0 root handle 1: \
red limit 500K avpkt 1K qevent early_drop block 10
# tc filter add block 10 \
matchall action mirred egress mirror dev eth1
This patch set introduces the corresponding iproute2 support. Patch #1 adds
the new netlink attribute enumerators. Patch #2 adds a set of helpers to
implement qevents, and #3 adds a generic documentation to tc.8. Patch #4
then adds two new qevents to the RED qdisc: mark and early_drop.
====================
Signed-off-by: David Ahern <dsahern@kernel.org>
The "early_drop" qevent matches packets that have been early-dropped. The
"mark" qevent matches packets that have been ECN-marked.
Signed-off-by: Petr Machata <petrm@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Introduce a set of helpers to make it easy to add support for qevents into
qdisc.
The idea behind this is that qevent types will be generally reused between
qdiscs, rather than each having a completely idiosyncratic set of qevents.
The qevent module holds functions for parsing, dumping and formatting of
these common qevent types, and for dispatch to the appropriate set of
handlers based on the qevent name.
Signed-off-by: Petr Machata <petrm@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Current police action must set 'rate' and 'burst'. 'mtu' parameter
set the max frame size and could be set alone without 'rate' and 'burst'
in some situation. Offloading to hardware for example, 'mtu' could limit
the flow max frame size.
Signed-off-by: Po Liu <po.liu@nxp.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Change the double quotes to single quotes in fprintf message to make it
more readable.
Signed-off-by: Po Liu <po.liu@nxp.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Following inclusion in net-next, extend rtnl_rtprot_tab and rt_protos
to support Keepalived.
Signed-off-by: Alexandre Cassen <acassen@gmail.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Parav Pandit says:
====================
Currently ip link set dev <pfndev> vf <vf_num> <param> <value> has
few below limitations.
1. Command is limited to set VF parameters only.
It cannot set the default MAC address for the PCI PF.
2. It can be set only on system where PCI SR-IOV is supported.
In smartnic based system, eswitch of a NIC resides on a different
embedded cpu which has the VF and PF representors for the SR-IOV
support on a host system in which this smartnic is plugged-in.
3. It cannot setup the function attributes of sub-function described
in detail in comprehensive RFC [1] and [2].
This series covers the first small part to let user query and set MAC
address (hardware address) of a PCI PF/VF which is represented by
devlink port.
[1] https://lore.kernel.org/netdev/20200519092258.GF4655@nanopsycho/
[2] https://marc.info/?l=linux-netdev&m=158555928517777&w=2
====================
Signed-off-by: David Ahern <dsahern@kernel.org>
Support setting devlink port function hardware address.
Example of a PCI VF port which supports a port function:
Set hardware address of the VF's port function.
$ devlink port show pci/0000:06:00.0/2
pci/0000:06:00.0/2: type eth netdev enp6s0pf0vf1 flavour pcivf pfnum 0 vfnum 1
function:
hw_addr 00:00:00:00:00:00
$ devlink port function set pci/0000:06:00.0/2 hw_addr 00:11:22:33:44:55
$ devlink port show pci/0000:06:00.0/2
pci/0000:06:00.0/2: type eth netdev enp6s0pf0vf1 flavour pcivf pfnum 0 vfnum 1
function:
hw_addr 00:11:22:33:44:55
Signed-off-by: Parav Pandit <parav@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Add support to query the hardware address of function represented
by devlink port function.
Example of a PCI VF port which supports a port function:
$ devlink port show pci/0000:06:00.0/2
pci/0000:06:00.0/2: type eth netdev enp6s0pf0vf1 flavour pcivf pfnum 0 vfnum 1
function:
hw_addr 00:11:22:33:44:66
$ devlink port show pci/0000:06:00.0/2 -jp
{
"port": {
"pci/0000:06:00.0/2": {
"type": "eth",
"netdev": "enp6s0pf0vf1",
"flavour": "pcivf",
"pfnum": 0,
"vfnum": 1,
"function": {
"hw_addr": "00:11:22:33:44:66"
}
}
}
}
Signed-off-by: Parav Pandit <parav@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
To reuse print routines for port function in subsequent patch, move
print routine specific to devlink device at start of the file.
Signed-off-by: Parav Pandit <parav@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
On some distros, i.e. rhel 7.6, compilation fails with the following:
ipaddress.c: In function ‘lookup_flag_data_by_name’:
ipaddress.c:1260:2: error: ‘for’ loop initial declarations are only allowed in C99 mode
for (int i = 0; i < ARRAY_SIZE(ifa_flag_data); ++i) {
^
ipaddress.c:1260:2: note: use option -std=c99 or -std=gnu99 to compile your code
This commit fixes the single place needed for compilation to pass.
Fixes: 9d59c86e57 ("iproute2: ip addr: Organize flag properties structurally")
Signed-off-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Vladyslav Tarasiuk
David Ahern
Petr Machata
Paolo Abeni
Eyal Birger
Hoang Huu Le
Julien Fortin
Tony Ambardar
Andrea Claudi
Stephen Hemminger
Louis Peens
Ido Schimmel
Guillaume Nault
Dmitry Yakunin
Sorah Fukumori
Anton Danilov
Bjarni Ingi Gislason
Maor Gottlieb
Po Liu
Alexandre Cassen
Parav Pandit
Roi Dayan