tc-mqprio(8) — Linux manual page
MQPRIO(8) Linux MQPRIO(8)
NAME
MQPRIO - Multiqueue Priority Qdisc (Offloaded Hardware QOS)
SYNOPSIS
tc qdisc ... dev dev ( parent classid | root) [ handle major: ]
mqprio
[ num_tc tcs ] [ map P0 P1 P2... ] [ queues
count1@offset1 count2@offset2 ... ]
[ hw 1|0 ] [ mode dcb|channel ] [ shaper dcb|bw_rlimit ]
[ min_rate min_rate1 min_rate2 ... ] [ max_rate max_rate1
max_rate2 ... ]
[ fp FP0 FP1 FP2 ... ]
DESCRIPTION
The MQPRIO qdisc is a simple queuing discipline that allows
mapping traffic flows to hardware queue ranges using priorities
and a configurable priority to traffic class mapping. A traffic
class in this context is a set of contiguous qdisc classes which
map 1:1 to a set of hardware exposed queues.
By default the qdisc allocates a pfifo qdisc (packet limited
first in, first out queue) per TX queue exposed by the lower
layer device. Other queuing disciplines may be added
subsequently. Packets are enqueued using the map parameter and
hashed across the indicated queues in the offset and count. By
default these parameters are configured by the hardware driver to
match the hardware QOS structures.
Channel mode supports full offload of the mqprio options, the
traffic classes, the queue configurations and QOS attributes to
the hardware. Enabled hardware can provide hardware QOS with the
ability to steer traffic flows to designated traffic classes
provided by this qdisc. Hardware based QOS is configured using
the shaper parameter. bw_rlimit with minimum and maximum
bandwidth rates can be used for setting transmission rates on
each traffic class. Also further qdiscs may be added to the
classes of MQPRIO to create more complex configurations.
ALGORITHM
On creation with 'tc qdisc add', eight traffic classes are
created mapping priorities 0..7 to traffic classes 0..7 and
priorities greater than 7 to traffic class 0. This requires base
driver support and the creation will fail on devices that do not
support hardware QOS schemes.
These defaults can be overridden using the qdisc parameters.
Providing the 'hw 0' flag allows software to run without hardware
coordination.
If hardware coordination is being used and arguments are provided
that the hardware can not support then an error is returned. For
many users hardware defaults should work reasonably well.
As one specific example numerous Ethernet cards support the
802.1Q link strict priority transmission selection algorithm
(TSA). MQPRIO enabled hardware in conjunction with the
classification methods below can provide hardware offloaded
support for this TSA.
CLASSIFICATION
Multiple methods are available to set the SKB priority which
MQPRIO uses to select which traffic class to enqueue the packet.
From user space
A process with sufficient privileges can encode the
destination class directly with SO_PRIORITY, see
socket(7).
with iptables/nftables
An iptables/nftables rule can be created to match traffic
flows and set the priority. iptables(8)
with net_prio cgroups
The net_prio cgroup can be used to set the priority of all
sockets belong to an application. See kernel and cgroup
documentation for details.
QDISC PARAMETERS
num_tc Number of traffic classes to use. Up to 16 classes
supported. You cannot have more classes than queues
map The priority to traffic class map. Maps priorities 0..15
to a specified traffic class.
queues Provide count and offset of queue range for each traffic
class. In the format, count@offset. Queue ranges for each
traffic classes cannot overlap and must be a contiguous
range of queues.
hw Set to 1 to support hardware offload. Set to 0 to
configure user specified values in software only. The
default value of this parameter is 1
mode Set to channel for full use of the mqprio options. Use dcb
to offload only TC values and use hardware QOS defaults.
Supported with 'hw' set to 1 only.
shaper Use bw_rlimit to set bandwidth rate limits for a traffic
class. Use dcb for hardware QOS defaults. Supported with
'hw' set to 1 only.
min_rate
Minimum value of bandwidth rate limit for a traffic class.
Supported only when the 'shaper' argument is set to
'bw_rlimit'.
max_rate
Maximum value of bandwidth rate limit for a traffic class.
Supported only when the 'shaper' argument is set to
'bw_rlimit'.
fp Selects whether traffic classes are express (deliver
packets via the eMAC) or preemptible (deliver packets via
the pMAC), according to IEEE 802.1Q-2018 clause 6.7.2
Frame preemption. Takes the form of an array (one element
per traffic class) with values being 'E' (for express) or
'P' (for preemptible).
Multiple priorities which map to the same traffic class,
as well as multiple TXQs which map to the same traffic
class, must have the same FP attributes. To interpret the
FP as an attribute per priority, the 'map' argument can be
used for translation. To interpret FP as an attribute per
TXQ, the 'queues' argument can be used for translation.
Traffic classes are express by default. The argument is
supported only with 'hw' set to 1. Preemptible traffic
classes are accepted only if the device has a MAC Merge
layer configurable through ethtool(8).
SEE ALSO
ethtool(8)
EXAMPLE
The following example shows how to attach priorities to 4 traffic
classes ("num_tc 4"), and then how to pair these traffic classes
with 4 hardware queues with mqprio, with hardware coordination
("hw 1", or does not specified, because 1 is the default value).
Traffic class 0 (tc0) is mapped to hardware queue 0 (q0), tc1 is
mapped to q1, tc2 is mapped to q2, and tc3 is mapped q3.
# tc qdisc add dev eth0 root mqprio num_tc 4 map 0 0 0 0 1 1 1 1 2 2 2 2 3 3 3 3 queues 1@0 1@1 1@2 1@3 hw 1
The next example shows how to attach priorities to 3 traffic
classes ("num_tc 3"), and how to pair these traffic classes with
4 queues, without hardware coordination ("hw 0"). Traffic class
0 (tc0) is mapped to hardware queue 0 (q0), tc1 is mapped to q1,
tc2 and is mapped to q2 and q3, where the queue selection between
these two queues is somewhat randomly decided.
# tc qdisc add dev eth0 root mqprio num_tc 3 map 0 0 0 0 1 1 1 1 2 2 2 2 2 2 2 2 queues 1@0 1@1 2@2 hw 0
In both cases from above the priority values from 0 to 3
(prio0-3) are mapped to tc0, prio4-7 are mapped to tc1, and the
prio8-11 are mapped to tc2 ("map" attribute). The last four
priority values (prio12-15) are mapped in different ways in the
two examples. They are mapped to tc3 in the first example and
mapped to tc2 in the second example. The values of these two
examples are the following:
┌────┬────┬───────┐ ┌────┬────┬────────┐
│Prio│ tc │ queue │ │Prio│ tc │ queue │
├────┼────┼───────┤ ├────┼────┼────────┤
│ 0 │ 0 │ 0 │ │ 0 │ 0 │ 0 │
│ 1 │ 0 │ 0 │ │ 1 │ 0 │ 0 │
│ 2 │ 0 │ 0 │ │ 2 │ 0 │ 0 │
│ 3 │ 0 │ 0 │ │ 3 │ 0 │ 0 │
│ 4 │ 1 │ 1 │ │ 4 │ 1 │ 1 │
│ 5 │ 1 │ 1 │ │ 5 │ 1 │ 1 │
│ 6 │ 1 │ 1 │ │ 6 │ 1 │ 1 │
│ 7 │ 1 │ 1 │ │ 7 │ 1 │ 1 │
│ 8 │ 2 │ 2 │ │ 8 │ 2 │ 2 or 3 │
│ 9 │ 2 │ 2 │ │ 9 │ 2 │ 2 or 3 │
│ 10 │ 2 │ 2 │ │ 10 │ 2 │ 2 or 3 │
│ 11 │ 2 │ 2 │ │ 11 │ 2 │ 2 or 3 │
│ 12 │ 3 │ 3 │ │ 12 │ 2 │ 2 or 3 │
│ 13 │ 3 │ 3 │ │ 13 │ 2 │ 2 or 3 │
│ 14 │ 3 │ 3 │ │ 14 │ 2 │ 2 or 3 │
│ 15 │ 3 │ 3 │ │ 15 │ 2 │ 2 or 3 │
└────┴────┴───────┘ └────┴────┴────────┘
example1 example2
Another example of queue mapping is the following. There are 5
traffic classes, and there are 8 hardware queues.
# tc qdisc add dev eth0 root mqprio num_tc 5 map 0 0 0 1 1 1 1 2 2 3 3 4 4 4 4 4 queues 1@0 2@1 1@3 1@4 3@5
The value mapping is the following for this example:
┌───────┐
tc0────┤Queue 0│◄────1@0
├───────┤
┌─┤Queue 1│◄────2@1
tc1──┤ ├───────┤
└─┤Queue 2│
├───────┤
tc2────┤Queue 3│◄────1@3
├───────┤
tc3────┤Queue 4│◄────1@4
├───────┤
┌─┤Queue 5│◄────3@5
│ ├───────┤
tc4──┼─┤Queue 6│
│ ├───────┤
└─┤Queue 7│
└───────┘
AUTHORS
John Fastabend, <john.r.fastabend@intel.com>
COLOPHON
This page is part of the iproute2 (utilities for controlling
TCP/IP networking and traffic) project. Information about the
project can be found at
⟨http://www.linuxfoundation.org/collaborate/workgroups/networking/iproute2⟩.
If you have a bug report for this manual page, send it to
netdev@vger.kernel.org, shemminger@osdl.org. This page was
obtained from the project's upstream Git repository
⟨https://git.kernel.org/pub/scm/network/iproute2/iproute2.git⟩ on
2024-06-14. (At that time, the date of the most recent commit
that was found in the repository was 2024-06-11.) If you
discover any rendering problems in this HTML version of the page,
or you believe there is a better or more up-to-date source for
the page, or you have corrections or improvements to the
information in this COLOPHON (which is not part of the original
manual page), send a mail to man-pages@man7.org
iproute2 24 Sept 2013 MQPRIO(8)
Pages that refer to this page: tc(8), tc-taprio(8)