1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2018 Intel Corporation.
7 The ice PMD (librte_pmd_ice) provides poll mode driver support for
8 10/25/50/100 Gbps IntelĀ® Ethernet 810 Series Network Adapters based on
9 the Intel Ethernet Controller E810.
15 - The E810 is currently in sampling state only. To obtain early samples and/or get further information
16 about kernel drivers, firmware and DDP support, please speak to your Intel representative.
18 - Follow the DPDK :ref:`Getting Started Guide for Linux <linux_gsg>` to setup the basic DPDK environment.
20 - To get better performance on Intel platforms, please follow the "How to get best performance with NICs on Intel platforms"
21 section of the :ref:`Getting Started Guide for Linux <linux_gsg>`.
23 Pre-Installation Configuration
24 ------------------------------
29 The following options can be modified in the ``config`` file.
30 Please note that enabling debugging options may affect system performance.
32 - ``CONFIG_RTE_LIBRTE_ICE_PMD`` (default ``y``)
34 Toggle compilation of the ``librte_pmd_ice`` driver.
36 - ``CONFIG_RTE_LIBRTE_ICE_DEBUG_*`` (default ``n``)
38 Toggle display of generic debugging messages.
40 - ``CONFIG_RTE_LIBRTE_ICE_16BYTE_RX_DESC`` (default ``n``)
42 Toggle to use a 16-byte RX descriptor, by default the RX descriptor is 32 byte.
44 Runtime Config Options
45 ~~~~~~~~~~~~~~~~~~~~~~
47 - ``Safe Mode Support`` (default ``0``)
49 If driver failed to load OS package, by default driver's initialization failed.
50 But if user intend to use the device without OS package, user can take ``devargs``
51 parameter ``safe-mode-support``, for example::
53 -w 80:00.0,safe-mode-support=1
55 Then the driver will be initialized successfully and the device will enter Safe Mode.
56 NOTE: In Safe mode, only very limited features are available, features like RSS,
57 checksum, fdir, tunneling ... are all disabled.
59 - ``Generic Flow Pipeline Mode Support`` (default ``0``)
61 In pipeline mode, a flow can be set at one specific stage by setting parameter
62 ``priority``. Currently, we support two stages: priority = 0 or !0. Flows with
63 priority 0 located at the first pipeline stage which typically be used as a firewall
64 to drop the packet on a blacklist(we called it permission stage). At this stage,
65 flow rules are created for the device's exact match engine: switch. Flows with priority
66 !0 located at the second stage, typically packets are classified here and be steered to
67 specific queue or queue group (we called it distribution stage), At this stage, flow
68 rules are created for device's flow director engine.
69 For none-pipeline mode, ``priority`` is ignored, a flow rule can be created as a flow director
70 rule or a switch rule depends on its pattern/action and the resource allocation situation,
71 all flows are virtually at the same pipeline stage.
72 By default, generic flow API is enabled in none-pipeline mode, user can choose to
73 use pipeline mode by setting ``devargs`` parameter ``pipeline-mode-support``,
76 -w 80:00.0,pipeline-mode-support=1
78 - ``Flow Mark Support`` (default ``0``)
80 This is a hint to the driver to select the data path that supports flow mark extraction
82 NOTE: This is an experimental devarg, it will be removed when any of below conditions
84 1) all data paths support flow mark (currently vPMD does not)
85 2) a new offload like RTE_DEV_RX_OFFLOAD_FLOW_MARK be introduced as a standard way to hint.
88 -w 80:00.0,flow-mark-support=1
90 - ``Protocol extraction for per queue``
92 Configure the RX queues to do protocol extraction into mbuf for protocol
93 handling acceleration, like checking the TCP SYN packets quickly.
95 The argument format is::
97 -w 18:00.0,proto_xtr=<queues:protocol>[<queues:protocol>...]
98 -w 18:00.0,proto_xtr=<protocol>
100 Queues are grouped by ``(`` and ``)`` within the group. The ``-`` character
101 is used as a range separator and ``,`` is used as a single number separator.
102 The grouping ``()`` can be omitted for single element group. If no queues are
103 specified, PMD will use this protocol extraction type for all queues.
105 Protocol is : ``vlan, ipv4, ipv6, ipv6_flow, tcp``.
107 .. code-block:: console
109 testpmd -w 18:00.0,proto_xtr='[(1,2-3,8-9):tcp,10-13:vlan]'
111 This setting means queues 1, 2-3, 8-9 are TCP extraction, queues 10-13 are
112 VLAN extraction, other queues run with no protocol extraction.
114 .. code-block:: console
116 testpmd -w 18:00.0,proto_xtr=vlan,proto_xtr='[(1,2-3,8-9):tcp,10-23:ipv6]'
118 This setting means queues 1, 2-3, 8-9 are TCP extraction, queues 10-23 are
119 IPv6 extraction, other queues use the default VLAN extraction.
121 The extraction metadata is copied into the registered dynamic mbuf field, and
122 the related dynamic mbuf flags is set.
124 .. table:: Protocol extraction : ``vlan``
126 +----------------------------+----------------------------+
128 +======+===+=================+======+===+=================+
129 | PCP | D | VID | PCP | D | VID |
130 +------+---+-----------------+------+---+-----------------+
132 VLAN1 - single or EVLAN (first for QinQ).
134 VLAN2 - C-VLAN (second for QinQ).
136 .. table:: Protocol extraction : ``ipv4``
138 +----------------------------+----------------------------+
140 +======+=======+=============+==============+=============+
141 | Ver |Hdr Len| ToS | TTL | Protocol |
142 +------+-------+-------------+--------------+-------------+
144 IPHDR1 - IPv4 header word 4, "TTL" and "Protocol" fields.
146 IPHDR2 - IPv4 header word 0, "Ver", "Hdr Len" and "Type of Service" fields.
148 .. table:: Protocol extraction : ``ipv6``
150 +----------------------------+----------------------------+
152 +=====+=============+========+=============+==============+
153 | Ver |Traffic class| Flow | Next Header | Hop Limit |
154 +-----+-------------+--------+-------------+--------------+
156 IPHDR1 - IPv6 header word 3, "Next Header" and "Hop Limit" fields.
158 IPHDR2 - IPv6 header word 0, "Ver", "Traffic class" and high 4 bits of
161 .. table:: Protocol extraction : ``ipv6_flow``
163 +----------------------------+----------------------------+
165 +=====+=============+========+============================+
166 | Ver |Traffic class| Flow Label |
167 +-----+-------------+-------------------------------------+
169 IPHDR1 - IPv6 header word 1, 16 low bits of the "Flow Label" field.
171 IPHDR2 - IPv6 header word 0, "Ver", "Traffic class" and high 4 bits of
174 .. table:: Protocol extraction : ``tcp``
176 +----------------------------+----------------------------+
177 | TCPHDR2 | TCPHDR1 |
178 +============================+======+======+==============+
179 | Reserved |Offset| RSV | Flags |
180 +----------------------------+------+------+--------------+
182 TCPHDR1 - TCP header word 6, "Data Offset" and "Flags" fields.
186 Use ``rte_net_ice_dynf_proto_xtr_metadata_get`` to access the protocol
187 extraction metadata, and use ``RTE_PKT_RX_DYNF_PROTO_XTR_*`` to get the
188 metadata type of ``struct rte_mbuf::ol_flags``.
190 The ``rte_net_ice_dump_proto_xtr_metadata`` routine shows how to
191 access the protocol extraction result in ``struct rte_mbuf``.
193 Driver compilation and testing
194 ------------------------------
196 Refer to the document :ref:`compiling and testing a PMD for a NIC <pmd_build_and_test>`
205 Vector PMD for RX and TX path are selected automatically. The paths
206 are chosen based on 2 conditions.
209 On the X86 platform, the driver checks if the CPU supports AVX2.
210 If it's supported, AVX2 paths will be chosen. If not, SSE is chosen.
212 - ``Offload features``
213 The supported HW offload features are described in the document ice_vec.ini.
214 If any not supported features are used, ICE vector PMD is disabled and the
215 normal paths are chosen.
217 Malicious driver detection (MDD)
218 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
220 It's not appropriate to send a packet, if this packet's destination MAC address
221 is just this port's MAC address. If SW tries to send such packets, HW will
222 report a MDD event and drop the packets.
224 The APPs based on DPDK should avoid providing such packets.
226 Device Config Function (DCF)
227 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
229 This section demonstrates ICE DCF PMD, which shares the core module with ICE
232 A DCF (Device Config Function) PMD bounds to the device's trusted VF with ID 0,
233 it can act as a sole controlling entity to exercise advance functionality (such
234 as switch, ACL) for the rest VFs.
236 The DCF PMD needs to advertise and acquire DCF capability which allows DCF to
237 send AdminQ commands that it would like to execute over to the PF and receive
238 responses for the same from PF.
242 .. figure:: img/ice_dcf.*
244 DCF Communication flow.
248 echo 4 > /sys/bus/pci/devices/0000\:18\:00.0/sriov_numvfs
250 #. Enable the VF0 trust on::
252 ip link set dev enp24s0f0 vf 0 trust on
254 #. Bind the VF0, and run testpmd with 'cap=dcf' devarg::
256 testpmd -l 22-25 -n 4 -w 18:01.0,cap=dcf -- -i
258 #. Monitor the VF2 interface network traffic::
260 tcpdump -e -nn -i enp24s1f2
262 #. Create one flow to redirect the traffic to VF2 by DCF::
264 flow create 0 priority 0 ingress pattern eth / ipv4 src is 192.168.0.2 \
265 dst is 192.168.0.3 / end actions vf id 2 / end
267 #. Send the packet, and it should be displayed on tcpdump::
269 sendp(Ether(src='3c:fd:fe:aa:bb:78', dst='00:00:00:01:02:03')/IP(src=' \
270 192.168.0.2', dst="192.168.0.3")/TCP(flags='S')/Raw(load='XXXXXXXXXX'), \
271 iface="enp24s0f0", count=10)
273 Sample Application Notes
274 ------------------------
279 Vlan filter only works when Promiscuous mode is off.
281 To start ``testpmd``, and add vlan 10 to port 0:
283 .. code-block:: console
285 ./app/testpmd -l 0-15 -n 4 -- -i
288 testpmd> rx_vlan add 10 0
290 Limitations or Known issues
291 ---------------------------
293 The Intel E810 requires a programmable pipeline package be downloaded
294 by the driver to support normal operations. The E810 has a limited
295 functionality built in to allow PXE boot and other use cases, but the
296 driver must download a package file during the driver initialization
299 The default DDP package file name is ice.pkg. For a specific NIC, the
300 DDP package supposed to be loaded can have a filename: ice-xxxxxx.pkg,
301 where 'xxxxxx' is the 64-bit PCIe Device Serial Number of the NIC. For
302 example, if the NIC's device serial number is 00-CC-BB-FF-FF-AA-05-68,
303 the device-specific DDP package filename is ice-00ccbbffffaa0568.pkg
304 (in hex and all low case). During initialization, the driver searches
305 in the following paths in order: /lib/firmware/updates/intel/ice/ddp
306 and /lib/firmware/intel/ice/ddp. The corresponding device-specific DDP
307 package will be downloaded first if the file exists. If not, then the
308 driver tries to load the default package. The type of loaded package
309 is stored in ``ice_adapter->active_pkg_type``.
311 A symbolic link to the DDP package file is also ok. The same package
312 file is used by both the kernel driver and the DPDK PMD.
317 Ice code released is for evaluation only currently.