devlink: print missing params even if an unknown one is present
Print all of the missing parameters, also in the presence of unknown ones.
Take for example a correct command:
$ devlink resource set pci/0000:01:00.0 path /kvd/linear size 98304
And remove the "size" keyword:
$ devlink resource set pci/0000:01:00.0 path /kvd/linear 98304
That yields output:
Resource size expected.
Unknown option "98304"
Prior to the patch only the last line of output was present. And if user
would forgot also the "path" keyword, there will be additional line:
Resource path expected.
in the stderr.
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> Signed-off-by: Przemek Kitszel <przemyslaw.kitszel@intel.com> Reviewed-by: Michal Kubiak <michal.kubiak@intel.com> Signed-off-by: David Ahern <dsahern@kernel.org>
Mark Zhang [Thu, 4 Jul 2024 06:29:01 +0000 (09:29 +0300)]
rdma: Supports to add/delete a device with type SMI
This patch adds a new device attribute "type", as well as supports to
add and delete a rdma device with a specific type. This new device
provides a subset of functionalists defined in IBTA spec.
Currently only type "SMI" is supported: A SMI device provides SMI (QP0)
interface; This device and it's parent associates with the same HCA port
and shares the physical link, so when the parent doesn't support SMI,
It allows the subnet manager to configure the link.
This patch also supports to print device type and parent if any.
Examples:
$ rdma dev add smi1 type SMI parent ibp8s0f1
$ rdma dev show smi1
2: smi1: node_type ca fw 20.38.1002 node_guid 9803:9b03:009f:d5ef sys_image_guid 9803:9b03:009f:d5ee type smi parent ibp8s0f1
$ rdma dev del smi1
Signed-off-by: Mark Zhang <markzhang@nvidia.com> Acked-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: David Ahern <dsahern@kernel.org>
Dragan Simic [Tue, 21 May 2024 10:04:52 +0000 (12:04 +0200)]
ss: use COLUMNS from the environment, if TIOCGWINSZ fails
Use the COLUMNS environment variable [1] when determining the screen width,
if using TIOCGWINSZ isn't possible or if it fails. This allows better use
of the available horizontal screen space in certain scenarios, and makes
the produced outputs more readable, as described further below.
All major shells can maintain the COLUMNS variable according to the current
screen size, [2][3][4] but this shell variable isn't actually an environment
variable, i.e. it doesn't get exported to the shell subprocesses by default.
For example, no COLUMNS environment variable reaches ss(8) when it's executed
as part of a shell pipeline or inside a shell script.
Though, users can opt to export the COLUMNS variable by hand, or they can
rely on some other utilities to do that for them. A good example of such
utilities is watch(1) that exports COLUMNS as an environment variable to
the processes it executes. [5] Using ss(8) together with watch(1) is rather
useful, and honoring the exported COLUMNS variable makes the outputs produced
by ss(8) in this scenario more readable.
The behavior of shells, which don't export the COLUMNS variable by default,
makes this change safe in the sense of not affecting the usual shell pipeline
workflows or various shell scripts that use ss(8).
Gabi Falk [Fri, 10 May 2024 14:36:12 +0000 (14:36 +0000)]
bridge/vlan.c: bridge/vlan.c: fix build with gcc 14 on musl systems
On glibc based systems the definition of 'struct timeval' is pulled in
with inclusion of <stdlib.h> header, but on musl based systems it
doesn't work this way. Missing definition triggers an
incompatible-pointer-types error with gcc 14 (warning on previous
versions of gcc):
../include/json_print.h:80:30: warning: 'struct timeval' declared inside parameter list will not be visible outside of this definition or declaration
80 | _PRINT_FUNC(tv, const struct timeval *)
| ^~~~~~~
../include/json_print.h:50:37: note: in definition of macro '_PRINT_FUNC'
50 | type value); \
| ^~~~
../include/json_print.h:80:30: warning: 'struct timeval' declared inside parameter list will not be visible outside of this definition or declaration
80 | _PRINT_FUNC(tv, const struct timeval *)
| ^~~~~~~
../include/json_print.h:55:45: note: in definition of macro '_PRINT_FUNC'
55 | type value) \
| ^~~~
../include/json_print.h: In function 'print_tv':
../include/json_print.h:58:48: error: passing argument 5 of 'print_color_tv' from incompatible pointer type [-Wincompatible-pointer-types]
58 | value); \
| ^~~~~
| |
| const struct timeval *
Signed-off-by: Gabi Falk <gabifalk@gmx.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
ip link: hsr: Add support for passing information about INTERLINK device
The HSR capable device can operate in two modes of operations -
Doubly Attached Node for HSR (DANH) and RedBOX (HSR-SAN).
The latter one allows connection of non-HSR aware device(s) to HSR
network.
This node is called SAN (Singly Attached Network) and is connected via
INTERLINK network device.
This patch adds support for passing information about the INTERLINK
device, so the Linux driver can properly setup it.
Signed-off-by: Lukasz Majewski <lukma@denx.de> Signed-off-by: David Ahern <dsahern@kernel.org>
rdma: Add an option to display driver-specific QPs in the rdma tool
Utilize the -dd flag (driver-specific details) in the rdmatool
to view driver-specific QPs which are not exposed yet.
The following examples show mlx5 UMR QP which is visible now:
$ rdma resource show qp link ibp8s0f1
link ibp8s0f1/1 lqpn 360 type UD state RTS sq-psn 0 comm [mlx5_ib]
link ibp8s0f1/1 lqpn 0 type SMI state RTS sq-psn 0 comm [ib_core]
link ibp8s0f1/1 lqpn 1 type GSI state RTS sq-psn 0 comm [ib_core]
$ rdma resource show qp link ibp8s0f1 -dd
link ibp8s0f1/1 lqpn 360 type UD state RTS sq-psn 0 comm [mlx5_ib]
link ibp8s0f1/1 lqpn 465 type DRIVER subtype REG_UMR state RTS sq-psn 0 comm [mlx5_ib]
link ibp8s0f1/1 lqpn 0 type SMI state RTS sq-psn 0 comm [ib_core]
link ibp8s0f1/1 lqpn 1 type GSI state RTS sq-psn 0 comm [ib_core]
If we forked, returning from the function will make the calling code to
continue in both the child and parent process. Make cmd_exec exit if
setup failed and it forked already.
An example of issues this causes, where a failure in setup causes
multiple unnecessary tries:
```
$ ip netns
ef
ab
$ ip -all netns exec ls
netns: ef
setting the network namespace "ef" failed: Operation not permitted
netns: ab
setting the network namespace "ab" failed: Operation not permitted
netns: ab
setting the network namespace "ab" failed: Operation not permitted
```
Signed-off-by: Yedaya Katsman <yedaya.ka@gmail.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
David Ahern [Tue, 23 Apr 2024 16:29:52 +0000 (16:29 +0000)]
Merge branch 'pfcp' into next
Wojciech Drewek says:
====================
New PFCP module was accepted in the kernel together with cls_flower
changes which allow to filter the packets using PFCP specific fields [1].
Packet Forwarding Control Protocol is a 3GPP Protocol defined in
TS 29.244 [2].
Extended ip link with the support for the new PFCP device.
Add pfcp_opts support in tc-flower.
Signed-off-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com> Signed-off-by: Wojciech Drewek <wojciech.drewek@intel.com> Signed-off-by: David Ahern <dsahern@kernel.org>
Wojciech Drewek [Mon, 22 Apr 2024 12:05:50 +0000 (14:05 +0200)]
ip: PFCP device support
Packet Forwarding Control Protocol is a 3GPP Protocol defined in
TS 29.244 [1]. Add support for PFCP device type in ip link.
It is capable of receiving PFCP messages and extracting its
metadata (session ID).
Its only purpose is to be used together with tc flower to create
SW/HW filters.
PFCP module does not take any netlink attributes so there is no
need to parse any args. Add new sections to the man to let the
user know about new device type.
man: use clsact qdisc for port mirroring examples on matchall and mirred
The clsact qdisc supports ingress and egress. Instead of using two qdiscs
to do ingress and egress port mirroring, clsact can be used. Therefore, use
clsact for the port mirroring examples on the tc-matchall.8 and tc-mirred.8
documents.
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
The version field in mnlu was being passed in but never set.
This meant that all places mnlu_gen_socket was used, the version would
be uninitialized data from malloc().
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Three new "last time" counters have been added to "struct mptcp_info":
last_data_sent, last_data_recv and last_ack_recv. They have been added
in commit 18d82cde7432 ("mptcp: add last time fields in mptcp_info") in
net-next recently.
This patch prints out these new counters into mptcp_stats output in ss.
Kernel has add IFLA_EXT_MASK attribute for indicating that certain
extended ifinfo values are requested by the user application. The ip
link show cmd always request VFs extended ifinfo.
In this case, RTM_GETLINK for greater than about 220 VFs truncates
IFLA_VFINFO_LIST due to the maximum reach of nlattr's nla_len being
exceeded. As a result, ip link show command only show the truncated
VFs info sucn as:
#ip link show dev eth0
1: eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 ...
link/ether ...
vf 0 link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff ...
Truncated VF list: eth0
This patch add novf to support filter links with no VF info:
ip link show novf
v2:
- use an one word option instead of an option with on/off.
- fix the issue that break changes made for the link filter
already done for VF's.
v3:
- "novf" set vfinfo to 0 and the RTEXT_FILTER_VF flag is not added.
Signed-off-by: Mingshuai Ren <renmingshuai@huawei.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: David Ahern <dsahern@kernel.org>
Max Gautier [Mon, 18 Mar 2024 15:49:13 +0000 (16:49 +0100)]
arpd: create /var/lib/arpd on first use
The motivation is to build distributions packages without /var to go
towards stateless systems, see link below (TL;DR: provisionning anything
outside of /usr on boot).
We only try do create the database directory when it's in the default
location, and assume its parent (/var/lib in the usual case) exists.
Links: https://0pointer.net/blog/projects/stateless.html Signed-off-by: Max Gautier <mg@max.gautier.name> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Date Huang [Fri, 22 Mar 2024 12:39:22 +0000 (20:39 +0800)]
bridge: vlan: fix compressvlans usage
Fix the incorrect short opt for compressvlans and color
in usage
Signed-off-by: Date Huang <tjjh89017@hotmail.com> Acked-by: Nikolay Aleksandrov <razor@blackwall.org> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
David Ahern [Fri, 15 Mar 2024 15:05:23 +0000 (15:05 +0000)]
Merge branch 'nexthop-grp-stats' into next
Petr Machata says:
====================
Next hop group stats allow verification of balancedness of a next hop
group. The feature was merged in kernel commit 7cf497e5a122 ("Merge branch
'nexthop-group-stats'"). This patchset adds to ip the corresponding
support.
NH group stats come in two flavors: as statistics for SW and for HW
datapaths. The former is shown when -s is given to "ip nexthop". The latter
demands more work from the kernel, and possibly driver and HW, and might
not be always necessary. Therefore tie it to -s -s, similarly to how ip
link shows more detailed stats when -s is given twice.
Here's an example usage:
# ip link add name gre1 up type gre \
local 172.16.1.1 remote 172.16.1.2 tos inherit
# ip nexthop replace id 1001 dev gre1
# ip nexthop replace id 1002 dev gre1
# ip nexthop replace id 1111 group 1001/1002 hw_stats on
# ip -s -s -j -p nexthop show id 1111
[ {
[ ...snip... ]
"hw_stats": {
"enabled": true,
"used": true
},
"group_stats": [ {
"id": 1001,
"packets": 0,
"packets_hw": 0
},{
"id": 1002,
"packets": 0,
"packets_hw": 0
} ]
} ]
hw_stats.enabled shows whether hw_stats have been requested for the given
group. hw_stats.used shows whether any driver actually implemented the
counter. group_stats[].packets show the total stats, packets_hw only the
HW-datapath stats.
Petr Machata [Thu, 14 Mar 2024 14:52:15 +0000 (15:52 +0100)]
ip: ipnexthop: Allow toggling collection of nexthop group HW statistics
Besides SW datapath stats, the kernel also support collecting statistics
from HW datapath, for nexthop groups offloaded to HW. Since collection of
these statistics may consume HW resources, there is an interface to request
that the HW stats be recorded. Add this toggle to "ip nexthop".
Signed-off-by: Petr Machata <petrm@nvidia.com> Signed-off-by: David Ahern <dsahern@kernel.org>
Petr Machata [Thu, 14 Mar 2024 14:52:14 +0000 (15:52 +0100)]
ip: ipnexthop: Support dumping next hop group HW stats
Besides SW datapath stats, the kernel also support collecting statistics
from HW datapath, for nexthop groups offloaded to HW. Request that these be
collected when ip is given "-s -s", similarly to how "ip link" shows more
statistics in that case.
Besides the statistics themselves, also show whether the collection of HW
statistics was in fact requested, and whether any driver actually
implemented the request.
Signed-off-by: Petr Machata <petrm@nvidia.com> Signed-off-by: David Ahern <dsahern@kernel.org>
Petr Machata [Thu, 14 Mar 2024 14:52:13 +0000 (15:52 +0100)]
ip: ipnexthop: Support dumping next hop group stats
Next hop group stats allow verification of balancedness of a next hop
group. The feature was merged in kernel commit 7cf497e5a122 ("Merge branch
'nexthop-group-stats'"). Add to ip the corresponding support. The
statistics are requested if "ip nexthop" is started with -s.
Signed-off-by: Petr Machata <petrm@nvidia.com> Signed-off-by: David Ahern <dsahern@kernel.org>
Petr Machata [Thu, 14 Mar 2024 14:52:12 +0000 (15:52 +0100)]
libnetlink: Add rta_getattr_uint()
NLA_UINT attributes have a 4-byte payload if possible, and an 8-byte one if
necessary. Add a function to extract these. Since we need to dispatch on
length anyway, make the getter truly universal by supporting also u8 and
u16.
Signed-off-by: Petr Machata <petrm@nvidia.com> Signed-off-by: David Ahern <dsahern@kernel.org>
The removal of tick usage in netem, means that some of the
helper functions in tc are no longer used and can be safely removed.
Other functions can be made static.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
The current version of netem in iproute2 has a maximum of 4.3 seconds
because of scaled 32 bit clock values. Some users would like to be
able to use larger delays to emulate things like storage delays.
Since kernel version 4.15, netem qdisc had netlink parameters
to express wider range of delays in nanoseconds. But the iproute2
side was never updated to use them.
This does break compatibility with older kernels (4.14 and earlier).
With these out of support kernels, the latency/delay parameter
will end up being ignored.
Reported-by: Marc Blanchet <marc.blanchet@viagenie.ca> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Lars Ellenberg [Fri, 1 Mar 2024 12:33:24 +0000 (13:33 +0100)]
ss: fix output of MD5 signature keys configured on TCP sockets
da9cc6ab introduced printing of MD5 signature keys when found.
But when changing printf() to out() calls with 90351722,
the implicit printf call in print_escape_buf() was overlooked.
That results in a funny output in the first line:
"<all-your-tcp-signature-keys-concatenated>State"
and ambiguity as to which of those bytes belong to which socket.
Add a static void out_escape_buf() immediately before we use it.
da9cc6ab (ss: print MD5 signature keys configured on TCP sockets, 2017-10-06) 90351722 (ss: Replace printf() calls for "main" output by calls to helper, 2017-12-12)
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
David Ahern [Tue, 27 Feb 2024 04:08:34 +0000 (04:08 +0000)]
Merge branch 'ss-socket-local-storage' into next
Quentin Deslandes says:
====================
BPF allows programs to store socket-specific data using
BPF_MAP_TYPE_SK_STORAGE maps. The data is attached to the socket itself,
and Martin added INET_DIAG_REQ_SK_BPF_STORAGES, so it can be fetched
using the INET_DIAG mechanism.
Currently, ss doesn't request the socket-local data, this patch aims to
fix this.
The first patch requests the socket-local data for the requested map ID
(--bpf-map-id=) or all the maps (--bpf-maps). It then prints the map_id
in COL_EXT.
Patch #2 uses libbpf and BTF to pretty print the map's content, like
`bpftool map dump` would do.
Patch #3 updates ss' man page to explain new options.
While I think it makes sense for ss to provide the socket-local storage
content for the sockets, it's difficult to conciliate the column-based
output of ss and having readable socket-local data. Hence, the
socket-local data is printed in a readable fashion over multiple lines
under its socket statistics, independently of the column-based approach.
Here is an example of ss' output with --bpf-maps:
[...]
ESTAB 340116 0 [...]
map_id: 114 [
(struct my_sk_storage){
.field_hh = (char)3,
(union){
.a = (int)17,
.b = (int)17,
},
}
]
Changed this series to an RFC as the merging window for net-next is
closed.
Changes from v8:
* Remove usage of libbpf_bpf_map_type_str() which requires libbpf-1.0+
and provide very little added value (David).
* Use ENABLE_BPF_SKSTORAGE_SUPPORT to gate the BPF socket-local storage
support, instead of HAVE_LIBBPF. iproute2 depends on libbpf-0.1, but
this change needs libbpf-0.5+. If the requirements are not met, ss can
still be compiled and used without BPF socket-local storage support, but
a warning will be printed at compile time.
Changes from v7:
* Fix comment format and checkpatch warnings (Stephen, David).
* Replaced Co-authored-by with Co-developed-by + Signed-off-by for
Martin's contribution on patch #1 to follow checkpatch requirements,
with Martin's approval.
Changes from v6:
* Remove column dedicated to BPF socket-local storage (COL_SKSTOR),
use COL_EXT instead (Matthieu).
Changes from v5:
* Add support for --oneline when printing socket-local data.
* Use \t to indent instead of " " to be consistent with other columns.
* Removed Martin's ack on patch #2 due to amount of lines changed.
Changes from v4:
* Fix return code for 2 calls.
* Fix issue when inet_show_netlink() retries a request.
* BPF dump object is created in bpf_map_opts_load_info().
Changes from v3:
* Minor refactoring to reduce number of HAVE_LIBBF usage.
* Update ss' man page.
* btf_dump structure created to print the socket-local data is cached
in bpf_map_opts. Creation of the btf_dump structure is performed if
needed, before printing the data.
* If a map can't be pretty-printed, print its ID and a message instead
of skipping it.
* If show_all=true, send an empty message to the kernel to retrieve all
the maps (as Martin suggested).
Changes from v2:
* bpf_map_opts_is_enabled is not inline anymore.
* Add more #ifdef HAVE_LIBBPF to prevent compilation error if
libbpf support is disabled.
* Fix erroneous usage of args instead of _args in vout().
* Add missing btf__free() and close(fd).
Changes from v1:
* Remove the first patch from the series (fix) and submit it separately.
* Remove double allocation of struct rtattr.
* Close BPF map FDs on exit.
* If bpf_map_get_fd_by_id() fails with ENOENT, print an error message
and continue to the next map ID.
* Fix typo in new command line option documentation.
* Only use bpf_map_info.btf_value_type_id and ignore
bpf_map_info.btf_vmlinux_value_type_id (unused for socket-local storage).
* Use btf_dump__dump_type_data() instead of manually using BTF to
pretty-print socket-local storage data. This change alone divides the size
of the patch series by 2.
ss is able to print the map ID(s) for which a given socket has BPF
socket-local storage defined (using --bpf-maps or --bpf-map-id=). However,
the actual content of the map remains hidden.
This change aims to pretty-print the socket-local storage content following
the socket details, similar to what `bpftool map dump` would do. The exact
output format is inspired by drgn, while the BTF data processing is similar
to bpftool's.
ss will use libbpf's btf_dump__dump_type_data() to ease pretty-printing
of binary data. This requires out_bpf_sk_storage_print_fn() as a print
callback function used by btf_dump__dump_type_data(). vout() is also
introduced, which is similar to out() but accepts a va_list as
parameter.
ss' output remains unchanged unless --bpf-maps or --bpf-map-id= is used,
in which case each socket containing BPF local storage will be followed by
the content of the storage before the next socket's info is displayed.
Signed-off-by: Quentin Deslandes <qde@naccy.de> Acked-by: Martin KaFai Lau <martin.lau@kernel.org> Signed-off-by: David Ahern <dsahern@kernel.org>
Yedaya Katsman [Sat, 17 Feb 2024 21:21:02 +0000 (23:21 +0200)]
ip: Add missing command exaplantions in man page
There are a few commands missing from the ip command syntax list, add
them. They are also missing from the see also section, add them there as
well.
Note there isn't a ip-ila man page, so I didn't link to it.
Also fix a few punctuation mistakes.
Signed-off-by: Yedaya Katsman <yedaya.ka@gmail.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
While sock_diag is able to return BPF socket-local storage in response
to INET_DIAG_REQ_SK_BPF_STORAGES requests, ss doesn't request it.
This change introduces the --bpf-maps and --bpf-map-id= options to request
BPF socket-local storage for all SK_STORAGE maps, or only specific ones.
The bigger part of this change will check the requested map IDs and
ensure they are valid. The column COL_EXT is used to print the
socket-local data into.
When --bpf-maps is used, ss will send an empty
INET_DIAG_REQ_SK_BPF_STORAGES request, in return the kernel will send
all the BPF socket-local storage entries for a given socket. The BTF
data for each map is loaded on demand, as ss can't predict which map ID
are used.
When --bpf-map-id=ID is used, a file descriptor to the requested maps is
open to 1) ensure the map doesn't disappear before the data is printed,
and 2) ensure the map type is BPF_MAP_TYPE_SK_STORAGE. The BTF data for
each requested map is loaded before the request is sent to the kernel.
Co-developed-by: Martin KaFai Lau <martin.lau@kernel.org> Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org> Signed-off-by: Quentin Deslandes <qde@naccy.de> Signed-off-by: David Ahern <dsahern@kernel.org>
Takanori Hirano [Sun, 11 Feb 2024 01:38:48 +0000 (01:38 +0000)]
tc: Change of json format in tc-fw
In the case of a process such as mapping a json to a structure,
it can be difficult if the keys have the same name but different types.
Since handle is used in hex string, change it to fw.
Signed-off-by: Takanori Hirano <me@hrntknr.net> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Throughout ifstat.c, ifstat_ent.val is accessed as a long long unsigned
type, however it is defined as __u64. This works by coincidence on many
systems, however on ppc64le, __u64 is a long unsigned.
This patch makes the type definition consistent with all of the places
where it is accessed.
Fixes: 5a52102b7c8f ("ifstat: Add extended statistics to ifstat") Reviewed-by: Andrea Claudi <aclaudi@redhat.com> Signed-off-by: Stephen Gallagher <sgallagh@redhat.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
projects / thirdparty / iproute2.git / log
