Change the code that is doing custom processing of boolean command line
options to use dl_argv_bool(). Extend strtobool() to accept
"enable"/"disable" to maintain current behavior.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Extend bash-completion for two new commands:
devlink trap policer set DEV policer POLICER [ rate RATE ] [ burst BURST ]
devlink trap policer show DEV policer POLICER
And for "policer" / "nopolicer" parameters in existing command:
devlink trap group set DEV group GROUP [ action { trap | drop } ]
[ policer POLICER ] [ nopolicer ]
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Add ability to associate a policer with a trap group. The policer can be
unbound by using the 'nopolicer' keyword. In which case, the value
encoded in the 'DEVLINK_ATTR_TRAP_POLICER_ID' attribute will be '0'.
This is consistent with ip-link 'nomaster' keyword and the 'IFLA_MASTER'
attribute.
Example:
# devlink trap group set netdevsim/netdevsim10 group l3_drops policer 2
# devlink -jp trap group show netdevsim/netdevsim10 group l3_drops
{
"trap_group": {
"netdevsim/netdevsim10": [ {
"name": "l3_drops",
"generic": true,
"policer": 2
} ]
}
}
# devlink trap group set netdevsim/netdevsim10 group l3_drops nopolicer
# devlink -jp trap group show netdevsim/netdevsim10 group l3_drops
{
"trap_group": {
"netdevsim/netdevsim10": [ {
"name": "l3_drops",
"generic": true
} ]
}
}
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
The trap policer set command allows the user to set the parameters of
the packet trap policer, such as rate and burst size. Example:
# devlink trap policer set netdevsim/netdevsim10 policer 1 rate 1000 burst 32
The trap policer show command allows the user to get the current
parameters of an individual policer or a dump of all policers in case
one is not specified. When '-s' is specified the policer's statistics
are shown. Example:
# devlink -jps trap policer show netdevsim/netdevsim10 policer 1
{
"trap_policer": {
"netdevsim/netdevsim10": [ {
"policer": 1,
"rate": 1000,
"burst": 32,
"stats": {
"rx": {
"dropped": 53
}
}
} ]
}
}
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
When multiple bridges are present, 'bridge link show' diplays ports
for all bridges. Make this clear in the command description, and
point out the user to the ip command to display ports for a specific
bridge.
Reported-by: Marc Muehlfeld <mmuehlfe@redhat.com>
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Add function for command completion for devlink in bash, and update Makefile
to install it under /usr/share/bash-completion/completions/.
Signed-off-by: Danielle Ratson <danieller@mellanox.com>
Tested-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Recognize the new configuration option of the RED Qdisc, "nodrop". Add
support for passing flags through TCA_RED_FLAGS, and use it when passing
TC_RED_NODROP flag.
Signed-off-by: Petr Machata <petrm@mellanox.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Follow the kernel rename to shorten the identifiers.
Rename hw_stats_type to hw_stats.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David Ahern <dsahern@gmail.com>
Introduce support for per-action hw stats type config.
This patch allows user to specify one of the following types of HW
stats for added action:
immediate - queried during dump time
delayed - polled from HW periodically or sent by HW in async manner
disabled - no stats needed
Note that if "hw_stats" option is not passed, user does not care about
the type, just expects any type of stats.
Examples:
$ tc filter add dev enp0s16np28 ingress proto ip handle 1 pref 1 flower skip_sw dst_ip 192.168.1.1 action drop hw_stats disabled
$ tc -s filter show dev enp0s16np28 ingress
filter protocol ip pref 1 flower chain 0
filter protocol ip pref 1 flower chain 0 handle 0x1
eth_type ipv4
dst_ip 192.168.1.1
skip_sw
in_hw in_hw_count 2
action order 1: gact action drop
random type none pass val 0
index 1 ref 1 bind 1 installed 7 sec used 2 sec
Action statistics:
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
hw_stats disabled
$ tc filter add dev enp0s16np28 ingress proto ip handle 1 pref 1 flower skip_sw dst_ip 192.168.1.1 action drop hw_stats immediate
$ tc -s filter show dev enp0s16np28 ingress
filter protocol ip pref 1 flower chain 0
filter protocol ip pref 1 flower chain 0 handle 0x1
eth_type ipv4
dst_ip 192.168.1.1
skip_sw
in_hw in_hw_count 2
action order 1: gact action drop
random type none pass val 0
index 1 ref 1 bind 1 installed 11 sec used 4 sec
Action statistics:
Sent 102 bytes 1 pkt (dropped 1, overlimits 0 requeues 0)
Sent software 0 bytes 0 pkt
Sent hardware 102 bytes 1 pkt
backlog 0b 0p requeues 0
hw_stats immediate
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
The initial value of "label" in parse_mpls() is 0xffffffff. Therefore
we should test for this value, and not 0, to detect if a label has been
provided. The "!label" test not only fails to detect a missing label
parameter, it also prevents the use of the IPv4 explicit NULL label,
which actually equals 0.
Reproducer:
$ ip link add name dm0 type dummy
$ tc qdisc add dev dm0 ingress
$ tc filter add dev dm0 parent ffff: matchall action mpls push
Error: act_mpls: Label is required for MPLS push.
We have an error talking to the kernel
--> Filter was pushed to the kernel, where it got rejected.
$ tc filter add dev dm0 parent ffff: matchall action mpls push label 0
Error: argument "label" is required
--> Label 0 was rejected by iproute2.
Expected result:
$ tc filter add dev dm0 parent ffff: matchall action mpls push
Error: argument "label" is required
--> Filter was directly rejected by iproute2.
$ tc filter add dev dm0 parent ffff: matchall action mpls push label 0
--> Filter is accepted.
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
nh_dump_filter is missing a return value check in two cases.
Fix this simply adding an assignment to the proper variable.
Fixes: 63df8e8543 ("Add support for nexthop objects")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
This reverts commit 92cfe3260e.
Kernel commit 3f95f55eb55d ("net: sched: pie: change tc_pie_xstats->prob")
removes the need to change the maximum integer value of
tc_pie_stats->prob here.
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Kernel commit 105e808c1da2 ("pie: remove pie_vars->accu_prob_overflows"),
changes the maximum value of tc_pie_xstats->prob from (2^64 - 1) to
(2^56 - 1).
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Antoine Tenart says:
====================
This series adds support for selecting and reporting the offloading mode
of a MACsec interface. Available modes are for now 'off' and 'phy',
'off' being the default when an interface is created. Modes are not only
'off' and 'on' as the MACsec operations can be offloaded to multiple
kinds of specialized hardware devices, at least to PHYs and Ethernet
MACs. The later isn't currently supported in the kernel though.
The first patch adds support for reporting the offloading mode currently
selected for a given MACsec interface through the `ip macsec show`
command:
# ip macsec show
18: macsec0: protect on validate strict sc off sa off encrypt on send_sci on end_station off scb off replay off
cipher suite: GCM-AES-128, using ICV length 16
TXSC: 3e5035b67c860001 on SA 0
0: PN 1, state on, key 00000000000000000000000000000000
RXSC: b4969112700f0001, state on
0: PN 1, state on, key 01000000000000000000000000000000
-> offload: phy
19: macsec1: protect on validate strict sc off sa off encrypt on send_sci on end_station off scb off replay off
cipher suite: GCM-AES-128, using ICV length 16
TXSC: 3e5035b67c880001 on SA 0
1: PN 1, state on, key 00000000000000000000000000000000
RXSC: b4969112700f0001, state on
1: PN 1, state on, key 01000000000000000000000000000000
-> offload: off
The second patch allows an user to change the offloading mode at runtime
through a new subcommand, `ip macsec offload`:
# ip macsec offload macsec0 phy
# ip macsec offload macsec0 off
If a mode isn't supported, `ip macsec offload` will report an issue
(-EOPNOTSUPP).
Giving the offloading mode when a macsec interface is created was
discussed; it is not implemented in this series. It could come later
on, when needed, as we'll still want to support updating the offloading
mode at runtime (what's implemented in this series).
====================
Signed-off-by: David Ahern <dsahern@gmail.com>
This patch adds an accessor for the validate_str array, to handle future
changes adding a member.
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Add a description of the `ip macsec offload` command used to select the
offloading mode on a macsec interface when the underlying device
supports it.
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
MacSEC can now be offloaded to specialized hardware devices. Offloading
is off by default when creating a new MACsec interface, but the mode can
be updated at runtime. This patch adds a new subcommand,
`ip macsec offload`, to allow users to select the offloading mode of a
MACsec interface. It takes the mode to switch to as an argument, which
can for now either be 'off' or 'phy':
# ip macsec offload macsec0 phy
# ip macsec offload macsec0 off
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
This patch adds support to report the MACsec offloading mode currently
being enabled, which as of now can either be 'off' or 'phy'. This
information is reported through the `ip macsec show` command:
# ip macsec show
18: macsec0: protect on validate strict sc off sa off encrypt on send_sci on end_station off scb off replay off
cipher suite: GCM-AES-128, using ICV length 16
TXSC: 3e5035b67c860001 on SA 0
0: PN 1, state on, key 00000000000000000000000000000000
RXSC: b4969112700f0001, state on
0: PN 1, state on, key 01000000000000000000000000000000
offload: phy
19: macsec1: protect on validate strict sc off sa off encrypt on send_sci on end_station off scb off replay off
cipher suite: GCM-AES-128, using ICV length 16
TXSC: 3e5035b67c880001 on SA 0
1: PN 1, state on, key 00000000000000000000000000000000
RXSC: b4969112700f0001, state on
1: PN 1, state on, key 01000000000000000000000000000000
offload: off
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Currently PCI PF and VF devlink devices register their ports as
physical port in non-representors mode.
Introduce a new port flavour as virtual so that virtual devices can
register 'virtual' flavour to make it more clear to users.
An example of one PCI PF and 2 PCI virtual functions, each having
one devlink port.
$ devlink port show
pci/0000:06:00.0/1: type eth netdev ens2f0 flavour physical port 0
pci/0000:06:00.2/1: type eth netdev ens2f2 flavour virtual port 0
pci/0000:06:00.3/1: type eth netdev ens2f3 flavour virtual port 0
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Flow action cookie has been recently added to kernel, print it out.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Add description for filters on rdma statistics show command.
Also add a filter description on the help message of the command.
Additionally, fix some whitespace issue in the man page.
Reported-by: Zhaojuan Guo <zguo@redhat.com>
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Add resource subcommand in the OBJECT section and a short
description for it.
Reported-by: Zhaojuan Guo <zguo@redhat.com>
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
In kernel space, ipcomp(sub) states used by main states are not
allowed to be deleted by users, they would be freed only when
all main states are destroyed and no one uses them.
In user space, ip xfrm sta deleteall doesn't filter these ipcomp
states out, and it causes errors:
# ip xfrm state add src 192.168.0.1 dst 192.168.0.2 spi 0x1000 \
proto comp comp deflate mode tunnel sel src 192.168.0.1 dst \
192.168.0.2 proto gre
# ip xfrm sta deleteall
Failed to send delete-all request
: Operation not permitted
This patch is to fix it by filtering ipcomp states with a check
xsinfo->id.proto == IPPROTO_IPIP.
Fixes: c7699875be ("Import patch ipxfrm-20040707_2.diff")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Add description to the vrf subcommand and a reference to the
dedicated man page.
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Currently `ip -6 route show` gives us this output:
sharpd@eva ~/i/ip (master)> ip -6 route show
::1 dev lo proto kernel metric 256 pref medium
4:5::6:7 nhid 18 proto static metric 20
nexthop via fe80::99 dev enp39s0 weight 1
nexthop via fe80::44 dev enp39s0 weight 1 pref medium
Displaying `pref medium` as the last bit of output implies
that the RTA_PREF is a per nexthop value, when it is infact
a per route piece of data.
Change the output to display RTA_PREF and RTA_TTL_PROPAGATE
before the RTA_MULTIPATH data is shown:
sharpd@eva ~/i/ip (master)> ./ip -6 route show
::1 dev lo proto kernel metric 256 pref medium
4:5::6:7 nhid 18 proto static metric 20 pref medium
nexthop via fe80::99 dev enp39s0 weight 1
nexthop via fe80::44 dev enp39s0 weight 1
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Andrea Claudi <aclaudi@redhat.com>
Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
When nstat temporary file is corrupted or in some other corner cases,
nstat use abort() to stop its execution. This can puzzle some users,
wondering what is the reason for the crash.
This commit replaces abort() with some meaningful error messages and exit()
Reported-by: Renaud Métrich <rmetrich@redhat.com>
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Commit 2897636267 ("erspan: add erspan version II support")
breaks the command:
# ip link add erspan1 type erspan key 1 seq erspan 123 \
local 10.1.0.2 remote 10.1.0.1
as erspan_ver is set to 0 by default, then IFLA_GRE_ERSPAN_INDEX
won't be set in gre_parse_opt().
# ip -d link show erspan1
...
erspan remote 10.1.0.1 local 10.1.0.2 ... erspan_index 0 erspan_ver 1
^^^^^^^^^^^^^^
This patch is to change to set erspan_ver to 1 by default.
Fixes: 2897636267 ("erspan: add erspan version II support")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: William Tu <u9012063@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Add support for devlink health error recovery status monitoring.
Update devlink-monitor man page accordingly.
Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
This patch adds support for the FQ-PIE packet Scheduler
Principles:
- Packets are classified on flows.
- This is a Stochastic model (as we use a hash, several flows might
be hashed to the same slot)
- Each flow has a PIE managed queue.
- Flows are linked onto two (Round Robin) lists,
so that new flows have priority on old ones.
- For a given flow, packets are not reordered.
- Drops during enqueue only.
- ECN capability is off by default.
- ECN threshold (if ECN is enabled) is at 10% by default.
- Uses timestamps to calculate queue delay by default.
Usage:
tc qdisc ... fq_pie [ limit PACKETS ] [ flows NUMBER ]
[ target TIME ] [ tupdate TIME ]
[ alpha NUMBER ] [ beta NUMBER ]
[ quantum BYTES ] [ memory_limit BYTES ]
[ ecn_prob PERCENTAGE ] [ [no]ecn ]
[ [no]bytemode ] [ [no_]dq_rate_estimator ]
defaults:
limit: 10240 packets, flows: 1024
target: 15 ms, tupdate: 15 ms (in jiffies)
alpha: 1/8, beta : 5/4
quantum: device MTU, memory_limit: 32 Mb
ecnprob: 10%, ecn: off
bytemode: off, dq_rate_estimator: off
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Sachin D. Patil <sdp.sachin@gmail.com>
Signed-off-by: V. Saicharan <vsaicharan1998@gmail.com>
Signed-off-by: Mohit Bhasi <mohitbhasi1998@gmail.com>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
This fixes broken tests in commit c4f5862994 ("ss: use compact output
for undetected screen width")
It also escapes stars as grep is used and more bugs could sneak under
the radar with the previous solution.
Signed-off-by: Peter Junos <petoju@gmail.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Warning seen on Ubuntu
devlink.c: In function ‘cmd_dev_flash’:
devlink.c:3071:3: warning: ignoring return value of ‘write’, declared with attribute warn_unused_result [-Wunused-result]
3071 | write(pipe_w, &err, sizeof(err));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Fixes: 9b13cddfe2 ("devlink: implement flash status monitoring")
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
This restore the string format we have before jsonification, adding a
missing space between v2 and v3 on TX IGMP reports string.
Fixes: a9bc23a792 ("ip: bridge: add xstats json support")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
After commit 8589eb4efd ("treewide: refactor help messages") help
messages for xfrm state and policy are broken, printing many times the
same protocol in UPSPEC section:
$ ip xfrm state help
[...]
UPSPEC := proto { { tcp | tcp | tcp | tcp } [ sport PORT ] [ dport PORT ] |
{ icmp | icmp | icmp } [ type NUMBER ] [ code NUMBER ] |
gre [ key { DOTTED-QUAD | NUMBER } ] | PROTO }
This happens because strxf_proto function is non-reentrant and gets called
multiple times in the same fprintf instruction.
This commit fix the issue avoiding calls to strxf_proto() with a constant
param, just hardcoding strings for protocol names.
Fixes: 8589eb4efd ("treewide: refactor help messages")
Signed-off-by: Andrea Claudi <aclaudi@redhat.com>
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
