Hangbin Liu [Mon, 23 Nov 2020 13:11:58 +0000 (21:11 +0800)]
lib: make ipvrf able to use libbpf and fix function name conflicts
There are directly calls in libbpf for bpf program load/attach.
So we could just use two wrapper functions for ipvrf and convert
them with libbpf support.
Function bpf_prog_load() is removed as it's conflict with libbpf
function name.
bpf.c is moved to bpf_legacy.c for later main libbpf support in
iproute2.
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com> Signed-off-by: Hangbin Liu <haliu@redhat.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Hangbin Liu [Mon, 23 Nov 2020 13:11:57 +0000 (21:11 +0800)]
iproute2: add check_libbpf() and get_libbpf_version()
This patch aim to add basic checking functions for later iproute2
libbpf support.
First we add check_libbpf() in configure to see if we have bpf library
support. By default the system libbpf will be used, but static linking
against a custom libbpf version can be achieved by passing libbpf DESTDIR
to variable LIBBPF_DIR for configure.
Another variable LIBBPF_FORCE is used to control whether to build iproute2
with libbpf. If set to on, then force to build with libbpf and exit if
not available. If set to off, then force to not build with libbpf.
When dynamically linking against libbpf, we can't be sure that the
version we discovered at compile time is actually the one we are
using at runtime. This can lead to hard-to-debug errors. So we add
a new file lib/bpf_glue.c and a helper function get_libbpf_version()
to get correct libbpf version at runtime.
Signed-off-by: Hangbin Liu <haliu@redhat.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Roi Dayan [Tue, 24 Nov 2020 12:28:10 +0000 (14:28 +0200)]
tc flower: fix parsing vlan_id and vlan_prio
When protocol is vlan then eth_type is set to the vlan eth type.
So when parsing vlan_id and vlan_prio need to check tc_proto
is vlan and not eth_type.
Fixes: 4c551369e083 ("tc flower: use right ethertype in icmp/arp parsing") Signed-off-by: Roi Dayan <roid@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Sat, 14 Nov 2020 22:53:59 +0000 (23:53 +0100)]
ip: iplink_bridge_slave: Convert to use print_on_off()
Instead of rolling a custom on-off printer, use the one added to utils.c.
Note that _print_onoff() has an extra parameter for a JSON-specific flag
name. However that argument is not used, and never was. Therefore when
moving over to print_on_off(), drop this argument.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Sat, 14 Nov 2020 22:53:58 +0000 (23:53 +0100)]
ip: iplink_bridge_slave: Port over to parse_on_off()
Invoke parse_on_off() from bridge_slave_parse_on_off() instead of
hand-rolling one. Exit on failure, because the invarg that was ivoked here
before would.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
David Ahern [Sun, 22 Nov 2020 19:46:01 +0000 (12:46 -0700)]
Merge branch 'nexthop-flags' into next
Ido Schimmel says:
====================
From: Ido Schimmel <idosch@nvidia.com>
Patch #1 prints the recently added 'RTNH_F_TRAP' flag.
Patch #2 makes sure that nexthop flags are always printed for nexthop
objects. Even when the nexthop does not have a device, such as a
blackhole nexthop or a group.
Example output with netdevsim:
$ ip nexthop
id 1 via 192.0.2.2 dev eth0 scope link trap
id 2 blackhole trap
id 3 group 2 trap
Example output with mlxsw:
$ ip nexthop
id 1 via 192.0.2.2 dev swp3 scope link offload
id 2 blackhole offload
id 3 group 2 offload
Tested with fib_nexthops.sh that uses "ip nexthop" output:
Ido Schimmel [Thu, 19 Nov 2020 13:57:31 +0000 (15:57 +0200)]
nexthop: Always print nexthop flags
Currently, the nexthop flags are only printed when the nexthop has a
nexthop device. The offload / trap indication is therefore not printed
for nexthop groups.
Instead, always print the nexthop flags, regardless if the nexthop has a
nexthop device or not.
Signed-off-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Ido Schimmel [Thu, 19 Nov 2020 13:57:30 +0000 (15:57 +0200)]
ip route: Print "trap" nexthop indication
The kernel can now signal that a nexthop is trapping packets instead of
forwarding them. Print the flag to help users understand the offload
state of each nexthop.
Signed-off-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Zahari Doychev [Tue, 10 Nov 2020 07:53:55 +0000 (08:53 +0100)]
tc flower: use right ethertype in icmp/arp parsing
Currently the icmp and arp parsing functions are called with incorrect
ethtype in case of vlan or cvlan filter options. In this case either
cvlan_ethtype or vlan_ethtype has to be used. The ethtype is now updated
each time a vlan ethtype is matched during parsing.
Signed-off-by: Zahari Doychev <zahari.doychev@linux.com> Signed-off-by: David Ahern <dsahern@gmail.com>
David Ahern [Sat, 14 Nov 2020 02:48:52 +0000 (19:48 -0700)]
Merge branch 'dcb-tool' into next
Petr Machata says:
====================
The Linux DCB interface allows configuration of a broad range of
hardware-specific attributes, such as TC scheduling, flow control, per-port
buffer configuration, TC rate, etc.
Currently a common libre tool for configuration of DCB is OpenLLDP. This
suite contains a daemon that uses Linux DCB interface to configure HW
according to the DCB TLVs exchanged over an interface. The daemon can also
be controlled by a client, through which the user can adjust and view the
configuration. The downside of using OpenLLDP is that it is somewhat
heavyweight and difficult to use in scripts, and does not support
extensions such as buffer and rate commands.
For access to many HW features, one would be perfectly fine with a
fire-and-forget tool along the lines of "ip" or "tc". For scripting in
particular, this would be ideal. This author is aware of one such tool,
mlnx_qos from Mellanox OFED scripts collection[1].
The downside here is that the tool is very verbose, the command line
language is awkward to use, it is not packaged in Linux distros, and
generally has the appearance of a very vendor-specific tool, despite not
being one.
This patchset addresses the above issues by providing a seed of a clean,
well-documented, easily usable, extensible fire-and-forget tool for DCB
configuration:
# dcb ets set dev eni1np1 \
tc-tsa all:strict 0:ets 1:ets 2:ets \
tc-bw all:0 0:33 1:33 2:34
- Many tools in iproute2 have an option to work in batch mode, where the
commands to run are given in a file. The code to handle batching is
largely the same independent of the tool in question. In patch #1, add a
helper to handle the batching, and migrate individual tools to use it.
- A number of configuration options come in a form of an on-off switch.
This in turn can be considered a special case of parsing one of a given
set of strings. In patch #2, extract helpers to parse one of a number of
strings, on top of which build an on-off parser.
Currently each tool open-codes the logic to parse the on-off toggle. A
future patch set will migrate instances of this code over to the new
helpers.
- The on/off toggles from previous list item sometimes need to be dumped.
While in the FP output, one typically wishes to maintain consistency with
the command line and show actual strings, "on" and "off", in JSON output
one would rather use booleans. This logic is somewhat annoying to have to
open-code time and again. Therefore in patch #3, add a helper to do just
that.
- The DCB tool is built on top of libmnl. Several routines will be
basically the same in DCB as they are currently in devlink. In patches
#4-#6, extract them to a new module, mnl_utils, for easy reuse.
- Much of DCB is built around arrays. A syntax similar to the iplink_vlan's
ingress-qos-map / egress-qos-map is very handy for describing changes
done to such arrays. Therefore in patch #7, extract a helper,
parse_mapping(), which manages parsing of key-value arrays. In patch #8,
fix a buglet in the helper, and in patch #9, extend it to allow setting
of all array elements in one go.
- In patch #10, add a skeleton of "dcb", which contains common helpers and
dispatches to subtools for handling of individual objects. The skeleton
is empty as of this patch.
In patch #11, add "dcb_ets", a module for handling of specifically DCB
ETS objects.
The intention is to gradually add handlers for at least PFC, APP, peer
configuration, buffers and rates.
Petr Machata [Thu, 12 Nov 2020 22:24:48 +0000 (23:24 +0100)]
dcb: Add a subtool for the DCB ETS object
ETS, for "Enhanced Transmission Selection", is a set of configurations that
permit configuration of mapping of priorities to traffic classes, traffic
selection algorithm to use per traffic class, bandwidth allocation, etc.
Add a dcb subtool to allow showing and tweaking of individual ETS
configuration options. For example:
Petr Machata [Thu, 12 Nov 2020 22:24:47 +0000 (23:24 +0100)]
Add skeleton of a new tool, dcb
The Linux DCB interface allows configuration of a broad range of
hardware-specific attributes, such as TC scheduling, flow control, per-port
buffer configuration, TC rate, etc. Add a new tool to show that
configuration and tweak it.
DCB allows configuration of several objects, and possibly could expand to
pre-standard CEE interfaces. Therefore the tool itself is a lean shell that
dispatches to subtools each dedicated to one of the objects.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Thu, 12 Nov 2020 22:24:46 +0000 (23:24 +0100)]
lib: parse_mapping: Recognize a keyword "all"
The DCB tool will have to provide an interface to a number of fixed-size
arrays. Unlike the egress- and ingress-qos-map, it makes good sense to have
an interface to set all members to the same value. For example to set
strict priority on all TCs besides select few, or to reset allocated
bandwidth to all zeroes, again besides several explicitly-given ones.
To support this usage, extend the parse_mapping() with a boolean that
determines whether this special use is supported. If "all" is given and
recognized, mapping_cb is called with the key of -1.
Have iplink_vlan pass false for allow_all.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Thu, 12 Nov 2020 22:24:45 +0000 (23:24 +0100)]
lib: parse_mapping: Update argc, argv on error
Currently argc and argv are not updated unless parsing of all of the
mapping was successful. However in that case, "ip link" will point at the
wrong argument when complaining:
# ip link add name eth0.100 link eth0 type vlan id 100 egress 1:1 2:foo
Error: argument "1" is wrong: invalid egress-qos-map
Update argc and argv even in the case of parsing error, so that the right
element is indicated.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Thu, 12 Nov 2020 22:24:44 +0000 (23:24 +0100)]
lib: Extract from iplink_vlan a helper to parse key:value arrays
VLAN netdevices have two similar attributes: ingress-qos-map and
egress-qos-map. These attributes can be configured with a series of
802.1-priority-to-skb-priority (and vice versa) mappings. A reusable helper
along those lines will be handy for configuration of various
priority-to-tc, tc-to-algorithm, and other arrays in DCB.
Therefore extract the logic to a function parse_mapping(), move to utils.c,
and dispatch to utils.c from iplink_vlan.c. That necessitates extraction of
a VLAN-specific parse_qos_mapping(). Do that, and propagate addattr_l()
return value up, unlike the original.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Thu, 12 Nov 2020 22:24:43 +0000 (23:24 +0100)]
lib: Extract from devlink/mnlg a helper, mnlu_socket_recv_run()
Receiving a message in libmnl is a somewhat involved operation. Devlink's
mnlg library has an implementation that is going to be handy for other
tools as well. Extract it into a new helper.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Thu, 12 Nov 2020 22:24:42 +0000 (23:24 +0100)]
lib: Extract from devlink/mnlg a helper, mnlu_msg_prepare()
Allocation of a new netlink message with the two usual headers is reusable
with other netlink netlink message types. Extract it into a helper,
mnlu_msg_prepare(). Take the second header as an argument, instead of
passing in parameters to initialize it, and copy it in.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Thu, 12 Nov 2020 22:24:41 +0000 (23:24 +0100)]
lib: Extract from devlink/mnlg a helper, mnlu_socket_open()
This little dance of mnl_socket_open(), option setting, and bind, is the
same regardless of tool. Extract into a new module that should hold helpers
for working with libmnl, mnl_util.c.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Thu, 12 Nov 2020 22:24:40 +0000 (23:24 +0100)]
lib: json_print: Add print_on_off()
The value of a number of booleans is shown as "on" and "off" in the plain
output, and as an actual boolean in JSON mode. Add a function that does
that.
RDMA tool already uses a function named print_on_off(). This function
always shows "on" and "off", even in JSON mode. Since there are probably
very few if any consumers of this interface at this point, migrate it to
the new central print_on_off() as well.
Signed-off-by: Petr Machata <me@pmachata.org> Reviewed-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Thu, 12 Nov 2020 22:24:39 +0000 (23:24 +0100)]
lib: Add parse_one_of(), parse_on_off()
Take from the macsec code parse_one_of() and adapt so that it passes the
primary result as the main return value, and error result through a
pointer. That is the simplest way to make the code reusable across data
types without introducing extra magic.
Also from macsec take the specialization of parse_one_of() for parsing
specifically the strings "off" and "on".
Convert the macsec code to the new helpers.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Petr Machata [Thu, 12 Nov 2020 22:24:38 +0000 (23:24 +0100)]
Unify batch processing across tools
The code for handling batches is largely the same across iproute2 tools.
Extract a helper to handle the batch, and adjust the tools to dispatch to
this helper. Sandwitch the invocation between prologue / epilogue code
specific for each tool.
Signed-off-by: Petr Machata <me@pmachata.org> Signed-off-by: David Ahern <dsahern@gmail.com>
David Ahern [Sat, 31 Oct 2020 15:18:43 +0000 (09:18 -0600)]
Merge branch 'tc-terse-dump' into next
Vlad Buslov says:
====================
Implement support for terse dump mode which provides only essential
classifier/action info (handle, stats, cookie, etc.). Use new
TCA_DUMP_FLAGS_TERSE flag to prevent copying of unnecessary data from
kernel.
Vlad Buslov [Fri, 23 Oct 2020 14:55:36 +0000 (17:55 +0300)]
tc: implement support for terse dump
Implement support for classifier/action terse dump using new TCA_DUMP_FLAGS
tlv with only available flag value TCA_DUMP_FLAGS_TERSE. Set the flag when
user requested it with following example CLI (-br for 'brief'):
$ tc -s -br filter show dev ens1f0 ingress
filter protocol ip pref 49151 flower chain 0
filter protocol ip pref 49151 flower chain 0 handle 0x1
not_in_hw
action order 1: gact Action statistics:
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
In terse mode dump only outputs essential data needed to identify the
filter and action (handle, cookie, etc.) and stats, if requested by the
user. The intention is to significantly improve rule dump rate by omitting
all static data that do not change after rule is created.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Vlad Buslov [Fri, 23 Oct 2020 14:55:35 +0000 (17:55 +0300)]
tc: skip actions that don't have options attribute when printing
Modify implementations that return error from action_until->print_aopt()
callback to silently skip actions that don't have their corresponding
TCA_ACT_OPTIONS attribute set (some actions already behave like this).
Print action kind before returning from action_until->print_aopt()
callbacks. This is necessary to support terse dump mode in following patch
in the series.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com> Suggested-by: Jamal Hadi Salim <jhs@mojatatu.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Johannes Berg [Mon, 26 Oct 2020 11:32:52 +0000 (12:32 +0100)]
libnetlink: define __aligned conditionally
On some systems (e.g. current Debian/stable) the inclusion
of utils.h pulled in some other things that may end up
defining __aligned, in a possibly different way than what
we had here.
Use our own definition only if there isn't one already.
Fixes: d5acae244f9d ("libnetlink: add nl_print_policy() helper") Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Guillaume Nault [Thu, 22 Oct 2020 09:11:44 +0000 (11:11 +0200)]
m_mpls: test the 'mac_push' action after 'modify'
Commit 02a261b5ba1c ("m_mpls: add mac_push action") added a matches()
test for the "mac_push" string before the test for "modify".
This changes the previous behaviour as 'action m' used to match
"modify" while it now matches "mac_push".
Revert to the original behaviour by moving the "mac_push" test after
"modify".
David Ahern [Tue, 20 Oct 2020 15:05:00 +0000 (09:05 -0600)]
Merge branch 'tipc-encryption' into next
Tuong Lien says:
====================
This series adds two new options in the 'iproute2/tipc' command, enabling users
to use the new TIPC encryption features, i.e. the master key and rekeying which
have been recently merged in kernel.
The help menu of the "tipc node set key" command is also updated accordingly:
# tipc node set key --help
Usage: tipc node set key KEY [algname ALGNAME] [PROPERTIES]
tipc node set key rekeying REKEYING
KEY
Symmetric KEY & SALT as a composite ASCII or hex string (0x...) in form:
[KEY: 16, 24 or 32 octets][SALT: 4 octets]
ALGNAME
Cipher algorithm [default: "gcm(aes)"]
PROPERTIES
master - Set KEY as a cluster master key
<empty> - Set KEY as a cluster key
nodeid NODEID - Set KEY as a per-node key for own or peer
REKEYING
INTERVAL - Set rekeying interval (in minutes) [0: disable]
now - Trigger one (first) rekeying immediately
EXAMPLES
tipc node set key this_is_a_master_key master
tipc node set key 0x746869735F69735F615F6B657931365F73616C74
tipc node set key this_is_a_key16_salt algname "gcm(aes)" nodeid 1001002
tipc node set key rekeying 600
Tuong Lien [Fri, 16 Oct 2020 16:02:01 +0000 (23:02 +0700)]
tipc: add option to set rekeying for encryption
As supported in kernel, the TIPC encryption rekeying can be tuned using
the netlink attribute - 'TIPC_NLA_NODE_REKEYING'. Now we add the
'rekeying' option correspondingly to the 'tipc node set key' command so
that user will be able to perform that tuning:
tipc node set key rekeying REKEYING
where the 'REKEYING' value can be:
INTERVAL - Set rekeying interval (in minutes) [0: disable]
now - Trigger one (first) rekeying immediately
For example:
$ tipc node set key rekeying 60
$ tipc node set key rekeying now
The command's help menu is also updated with these descriptions for the
new command option.
Acked-by: Jon Maloy <jmaloy@redhat.com> Signed-off-by: Tuong Lien <tuong.t.lien@dektech.com.au> Signed-off-by: David Ahern <dsahern@gmail.com>
Tuong Lien [Fri, 16 Oct 2020 16:02:00 +0000 (23:02 +0700)]
tipc: add option to set master key for encryption
In addition to the support of master key in kernel, we add the 'master'
option to the 'tipc node set key' command for user to be able to
specify a key as master key during the key setting. This is carried out
by turning on the new netlink flag - 'TIPC_NLA_NODE_KEY_MASTER'.
For example:
$ tipc node set key "this_is_a_master_key" master
The command's help menu is also updated to give a better description of
all the available options.
Acked-by: Jon Maloy <jmaloy@redhat.com> Signed-off-by: Tuong Lien <tuong.t.lien@dektech.com.au> Signed-off-by: David Ahern <dsahern@gmail.com>
David Ahern [Tue, 20 Oct 2020 14:57:47 +0000 (08:57 -0600)]
Merge branch 'tc-mpls-l2-vpn' into next
Guillaume Nault says:
====================
This patch series adds the possibility for TC to tunnel Ethernet frames
over MPLS.
Patch 1 allows adding or removing the Ethernet header.
Patch 2 allows pushing an MPLS LSE before the MAC header.
By combining these actions, it becomes possible to encapsulate an
entire Ethernet frame into MPLS, then add an outer Ethernet header
and send the resulting frame to the next hop.
Guillaume Nault [Mon, 19 Oct 2020 15:23:08 +0000 (17:23 +0200)]
m_mpls: add mac_push action
Add support for the new TCA_MPLS_ACT_MAC_PUSH action (kernel commit a45294af9e96 ("net/sched: act_mpls: Add action to push MPLS LSE before
Ethernet header")). This action let TC push an MPLS header before the
MAC header of a frame.
Example (encapsulate all outgoing frames with label 20, then add an
outer Ethernet header):
# tc filter add dev ethX matchall \
action mpls mac_push label 20 ttl 64 \
action vlan push_eth dst_mac 0a:00:00:00:00:02 \
src_mac 0a:00:00:00:00:01
This patch also adds an alias for ETH_P_TEB, since it is useful when
decapsulating MPLS packets that contain an Ethernet frame.
With MAC_PUSH, there's no previous Ethertype to modify. However, the
"protocol" option is still needed, because the kernel uses it to set
skb->protocol. So rename can_modify_ethtype() to can_set_ethtype().
Also add a test suite for m_mpls, which covers the new action and the
pre-existing ones.
Signed-off-by: Guillaume Nault <gnault@redhat.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Guillaume Nault [Mon, 19 Oct 2020 15:23:01 +0000 (17:23 +0200)]
m_vlan: add pop_eth and push_eth actions
Add support for the new TCA_VLAN_ACT_POP_ETH and TCA_VLAN_ACT_PUSH_ETH
actions (kernel commit 19fbcb36a39e ("net/sched: act_vlan:
Add {POP,PUSH}_ETH actions"). These action let TC remove or add the
Ethernet at the head of a frame.
Drop an Ethernet header:
# tc filter add dev ethX matchall action vlan pop_eth
Push an Ethernet header (the original frame must have no MAC header):
# tc filter add dev ethX matchall action vlan \
push_eth dst_mac 0a:00:00:00:00:02 src_mac 0a:00:00:00:00:01
Also add a test suite for m_vlan, which covers these new actions and
the pre-existing ones.
Signed-off-by: Guillaume Nault <gnault@redhat.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Jacob Keller [Wed, 14 Oct 2020 22:31:04 +0000 (15:31 -0700)]
devlink: display elapsed time during flash update
For some devices, updating the flash can take significant time during
operations where no status can meaningfully be reported. This can be
somewhat confusing to a user who sees devlink appear to hang on the
terminal waiting for the device to update.
Recent changes to the kernel interface allow such long running commands
to provide a timeout value indicating some upper bound on how long the
relevant action could take.
Provide a ticking counter of the time elapsed since the previous status
message in order to make it clear that the program is not simply stuck.
Display this message whenever the status message from the kernel
indicates a timeout value. Additionally also display the message if
we've received no status for more than couple of seconds. If we elapse
more than the timeout provided by the status message, replace the
timeout display with "timeout reached".
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> Reviewed-by: Jiri Pirko <jiri@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Jan Engelhardt [Mon, 12 Oct 2020 13:55:55 +0000 (15:55 +0200)]
ip: add error reporting when RTM_GETNSID failed
`ip addr` when run under qemu-user-riscv64, fails. This likely is due
to qemu-5.1 not doing translation of RTM_GETNSID calls. Aborting ip
completely is not helpful for the user however. This patch reworks
the error handling.
Before:
rtest:/ # ip a
2: host0@if4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
request send failed: Operation not supported
link/ether 46:3f:2d:88:3d:db brd ff:ff:ff:ff:ff:ffrtest:/ #
Afterwards:
rtest:/ # ip a
2: host0@if4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
rtnl_send(RTM_GETNSID): Operation not supported. Continuing anyway.
link/ether 46:3f:2d:88:3d:db brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet 192.168.72.147/28 brd 192.168.72.159 scope global host0
valid_lft forever preferred_lft forever
inet6 fe80::443f:2dff:fe88:3ddb/64 scope link
valid_lft forever preferred_lft forever
Signed-off-by: Jan Engelhardt <jengelh@inai.de> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
David Ahern [Mon, 12 Oct 2020 02:09:14 +0000 (20:09 -0600)]
Merge branch 'bridge-igmpv3-mldv2' into next
Nikolay Aleksandrov says:
====================
This set adds support for IGMPv3/MLDv2 attributes, they're mostly
read-only at the moment. The only new "set" option is the source address
for S,G entries. It is added in patch 01 (see the patch commit message for
an example). Patch 02 shows a missing flag (fast_leave) for
completeness, then patch 03 shows the new IGMPv3/MLDv2 flags:
added_by_star_ex and blocked. Patches 04-06 show the new extra
information about the entry's state when IGMPv3/MLDv2 are enabled. That
includes its filter mode (include/exclude), source list with timers and
origin protocol (currently only static/kernel), in order to show the new
information the user must use "-d"/show_details.
Here's the output of a few IGMPv3 entries:
dev bridge port ens12 grp 239.0.0.1 src 20.21.22.23 temp filter_mode include proto kernel blocked 0.00
dev bridge port ens12 grp 239.0.0.1 src 8.9.10.11 temp filter_mode include proto kernel blocked 0.00
dev bridge port ens12 grp 239.0.0.1 src 1.2.3.1 temp filter_mode include proto kernel blocked 0.00
dev bridge port ens12 grp 239.0.0.1 temp filter_mode exclude source_list 20.21.22.23/0.00,8.9.10.11/0.00,1.2.3.1/0.00 proto kernel 26.65
Print the mdb entry's protocol (i.e. who added it) when it's available if
the user requested to show details (-d). Currently the only possible
values are RTPROT_STATIC (user-space added) or RTPROT_KERNEL
(automatically added by kernel). The value is kernel controlled.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Print the mdb entry's source list when it's available if the user
requested to show details (-d). Each source has an associated timer
which controls if traffic should be forwarded to that S,G entry (if the
timer is non-zero traffic is forwarded, otherwise it's not).
Currently the source list is kernel controlled and can't be changed by
user-space.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Print the mdb entry's filter mode when it's available if the user
requested to show details (-d). It can be either include or exclude.
Currently it's kernel controlled and can't be changed by user-space.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
With IGMPv3/MLDv2 support we have 2 new flags:
- added_by_star_ex: set when the S,G entry was automatically created
because of a *,G entry in EXCLUDE mode
- blocked: set when traffic for the S,G entry for that port has to be
blocked
Both flags are used only on the new S,G entries and are currently kernel
managed, i.e. similar to other flags which can't be set from user-space.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
We're not showing the fast_leave flag when it's set. Currently that can
be only when an mdb entry is being deleted due to fast leave, so it will
only affect mdb monitor.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
This patch adds the user-space control and dump of mdb entry source
address. When setting the new MDBA_SET_ENTRY_ATTRS nested attribute is
used and inside is added MDBE_ATTR_SOURCE based on the address family.
When dumping we look for MDBA_MDB_EATTR_SOURCE and if present we add the
"src x.x.x.x" output. The source address will be always shown as it's
needed to match the entry to modify it from user-space.
Example:
$ bridge mdb add dev bridge port ens13 grp 239.0.0.1 src 1.2.3.4 permanent vid 100
$ bridge mdb show
dev bridge port ens13 grp 239.0.0.1 src 1.2.3.4 permanent vid 100
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Antony Antony [Fri, 2 Oct 2020 13:22:38 +0000 (15:22 +0200)]
ip xfrm: support setting XFRMA_SET_MARK_MASK attribute in states
The XFRMA_SET_MARK_MASK attribute can be set in states (4.19+)
It is optional and the kernel default is 0xffffffff
It is the mask of XFRMA_SET_MARK(a.k.a. XFRMA_OUTPUT_MARK in 4.18)
e.g.
./ip/ip xfrm state add output-mark 0x6 mask 0xab proto esp \
auth digest_null 0 enc cipher_null ''
ip xfrm state
src 0.0.0.0 dst 0.0.0.0
proto esp spi 0x00000000 reqid 0 mode transport
replay-window 0
output-mark 0x6/0xab
auth-trunc digest_null 0x30 0
enc ecb(cipher_null)
anti-replay context: seq 0x0, oseq 0x0, bitmap 0x00000000
sel src 0.0.0.0/0 dst 0.0.0.0/0
Signed-off-by: Antony Antony <antony@phenome.org> Signed-off-by: David Ahern <dsahern@gmail.com>
Jacob Keller [Wed, 30 Sep 2020 21:05:47 +0000 (14:05 -0700)]
devlink: support setting the overwrite mask attribute
The recently added DEVLINK_ATTR_FLASH_UPDATE_OVERWRITE_MASK allows
userspace to indicate how a device should handle subsections of a flash
component when updating. For example, a flash component might contain
vital data such as PCIe serial number or configuration fields such as
settings that control device bootup.
The overwrite mask allows specifying whether the device should overwrite
these subsections when updating from the provided image. If nothing is
specified, then the update is expected to preserve all vital fields and
configuration.
Add support for specifying the overwrite mask using the new "overwrite"
option to the flash command line.
By specifying "overwrite identifiers", the user request that the flash
update should overwrite any settings in the updated flash component with
settings from the provided flash image
$devlink dev flash pci/0000:af:00.0 file flash_image.bin overwrite identifiers
By specifying "overwrite settings" the user requests that the flash update
should overwrite any settings in the updated flash component with setting
values from the provided flash image.
$devlink dev flash pci/0000:af:00.0 file flash_image.bin overwrite settings
These options may be combined, in which case both subsections will be sent
in the overwrite mask, resulting in a request to overwrite all settings and
identifiers stored in the updated flash components.
$devlink dev flash pci/0000:af:00.0 file flash_image.bin overwrite settings overwrite identifiers
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> Signed-off-by: David Ahern <dsahern@gmail.com>
# ip address add 192.0.2.2/24 dev dummy5 noprefixroute
Warning: noprefixroute option can be set only for IPv6 addresses
# ip address add 224.1.1.10/24 dev dummy5 autojoin
Warning: autojoin option can be set only for IPv6 addresses
This enables them back for IPv4.
Fixes: 9d59c86e575b5 ("iproute2: ip addr: Organize flag properties structurally") Signed-off-by: Adel Belhouane <bugs.a.b@free.fr> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Kamal Heib [Sun, 27 Sep 2020 12:06:56 +0000 (15:06 +0300)]
ip: iplink_ipoib.c: Remove extra spaces
Remove the extra space between the reported ipoib attrs - use only one
space instead of two.
Fixes: de0389935f8c ("iplink: Added support for the kernel IPoIB RTNL ops") Signed-off-by: Kamal Heib <kamalheib1@gmail.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Jan Engelhardt [Mon, 28 Sep 2020 19:08:01 +0000 (21:08 +0200)]
build: avoid make jobserver warnings
I observe:
» make -j8 CCOPTS=-ggdb3
lib
make[1]: warning: -j8 forced in submake: resetting jobserver mode.
make[1]: Nothing to be done for 'all'.
ip
make[1]: warning: -j8 forced in submake: resetting jobserver mode.
CC ipntable.o
MFLAGS is a historic variable of some kind; removing it fixes the
jobserver issue.
Signed-off-by: Jan Engelhardt <jengelh@inai.de> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Plus those drivers are generally more modern than those
using rx_over_errors.
Since recently merged kernel documentation makes this
preference official, let's make ip -s output more informative
and let rx_missed_errors take the place of rx_over_errors.
David Ahern [Wed, 23 Sep 2020 02:17:48 +0000 (20:17 -0600)]
Merge branch 'devlink-controller-external-info' into next
Parav Pandit says:
====================
For certain devlink port flavours controller number and optionally external=
attributes are reported by the kernel.
(a) controller number indicates that a given port belong to which local or =
external controller.
(b) external port attribute indicates that if a given port is for external =
or local controller.
Wei Wang [Wed, 19 Aug 2020 21:13:54 +0000 (14:13 -0700)]
iproute2: ss: add support to expose various inet sockopts
This commit adds support to expose the following inet socket options:
-- recverr
-- is_icsk
-- freebind
-- hdrincl
-- mc_loop
-- transparent
-- mc_all
-- nodefrag
-- bind_address_no_port
-- recverr_rfc4884
-- defer_connect
with the option --inet-sockopt. The individual option is only shown
when set.
Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Hoang Le [Thu, 27 Aug 2020 02:30:37 +0000 (09:30 +0700)]
tipc: support 128bit node identity for peer removing
Problem:
In kernel upstream, we add the support to set node identity with
128bit. However, we are still using legacy format in command tipc
peer removing. Then, we got a problem when trying to remove
offline node i.e:
Roopa Prabhu [Sat, 29 Aug 2020 03:42:56 +0000 (20:42 -0700)]
iplink: add support for protodown reason
This patch adds support for recently
added link IFLA_PROTO_DOWN_REASON attribute.
IFLA_PROTO_DOWN_REASON enumerates reasons
for the already existing IFLA_PROTO_DOWN link
attribute.
$ ip link set dev vx10 protodown on protodown_reason vrrp on
$ip link show dev vx10
14: vx10: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT group default qlen 1000
link/ether f2:32:28:b8:35:ff brd ff:ff:ff:ff:ff:ff protodown on
protodown_reason <vrrp>
$ip -p -j link show dev vx10
[ {
<snip>
"proto_down": true,
"proto_down_reason": [ "vrrp" ]
} ]
$ip link set dev vx10 protodown_reason mlag on
$ip link show dev vx10
14: vx10: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT group default qlen 1000
link/ether f2:32:28:b8:35:ff brd ff:ff:ff:ff:ff:ff protodown on
protodown_reason <mlag,vrrp>
$ip -p -j link show dev vx10
[ {
<snip>
"proto_down": true,
"protodown_reason": [ "mlag","vrrp" ]
} ]
$ip -p -j link show dev vx10
$ip link set dev vx10 protodown off protodown_reason vrrp off
Error: Cannot clear protodown, active reasons.
$ip link set dev vx10 protodown off protodown_reason mlag off
$
Note: for somereason the json and non-json key for protodown
are different (protodown and proto_down). I have kept the
same for protodown reason for consistency (protodown_reason and
proto_down_reason).
Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Signed-off-by: David Ahern <dsahern@gmail.com>
Phil Sutter [Sat, 29 Aug 2020 10:18:35 +0000 (12:18 +0200)]
ip link: Fix indenting in help text
Indenting of 'ip link set' options below 'link-netns' was wrong, they
should be on the same level as the above.
While being at it, fix closing brackets in vf-specific options. Also
write node/port_guid parameters in upper-case without curly braces: They
are supposed to be replaced by values, not put literally.
Fixes: 8589eb4efdf2a ("treewide: refactor help messages") Fixes: 5a3ec4ba64783 ("iplink: Update usage in help message") Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Johannes Berg [Mon, 24 Aug 2020 17:51:08 +0000 (19:51 +0200)]
genl: ctrl: support dumping netlink policy
Support dumping the netlink policy of a given generic netlink
family, the policy (with any sub-policies if appropriate) is
exported by the kernel in a general fashion.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: David Ahern <dsahern@gmail.com>
Johannes Berg [Mon, 24 Aug 2020 17:51:07 +0000 (19:51 +0200)]
libnetlink: add nl_print_policy() helper
This prints out the data from the given nested attribute
to the given FILE pointer, interpreting the firmware that
the kernel has for showing netlink policies.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: David Ahern <dsahern@gmail.com>
Murali Karicheri [Mon, 17 Aug 2020 21:17:36 +0000 (17:17 -0400)]
iplink: hsr: add support for creating PRP device similar to HSR
This patch enhances the iplink command to add a proto parameters to
create PRP device/interface similar to HSR. Both protocols are
quite similar and requires a pair of Ethernet interfaces. So re-use
the existing HSR iplink command to create PRP device/interface as
well. Use proto parameter to differentiate the two protocols.
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Sascha Hauer [Mon, 17 Aug 2020 11:25:19 +0000 (13:25 +0200)]
iproute2: ip maddress: Check multiaddr length
ip maddress add|del takes a MAC address as argument, so insist on
getting a length of ETH_ALEN bytes. This makes sure the passed argument
is actually a MAC address and especially not an IPv4 address which
was previously accepted and silently taken as a MAC address.
While at it, do not print *argv in the error path as this has been
modified by ll_addr_a2n() and doesn't contain the full string anymore,
which can lead to misleading error messages.
Also while at it, replace the hardcoded buffer size with the actual
buffer size using sizeof().
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Leon Romanovsky [Tue, 11 Aug 2020 07:32:01 +0000 (10:32 +0300)]
rdma: Properly print device and link names in CLI output
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: b0a688a542cd ("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>
Leon Romanovsky [Tue, 11 Aug 2020 07:32:00 +0000 (10:32 +0300)]
rdma: Fix owner name for the kernel resources
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: b0a688a542cd ("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>
Mark Zhang [Tue, 4 Aug 2020 08:49:09 +0000 (11:49 +0300)]
rdma: Document the new "pid" criteria for auto mode
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>