xref: /dpdk/doc/guides/nics/ice.rst (revision d38febb08d57fec29fed27a2d12a507fc6fcdfa1)
1..  SPDX-License-Identifier: BSD-3-Clause
2    Copyright(c) 2018 Intel Corporation.
3
4ICE Poll Mode Driver
5======================
6
7The ice PMD (**librte_net_ice**) provides poll mode driver support for
810/25/50/100 Gbps Intel® Ethernet 800 Series Network Adapters based on
9the Intel Ethernet Controller E810 and Intel Ethernet Connection E822/E823.
10
11Linux Prerequisites
12-------------------
13
14- Follow the DPDK :ref:`Getting Started Guide for Linux <linux_gsg>` to setup the basic DPDK environment.
15
16- To get better performance on Intel platforms, please follow the "How to get best performance with NICs on Intel platforms"
17  section of the :ref:`Getting Started Guide for Linux <linux_gsg>`.
18
19- Please follow the matching list to download specific kernel driver, firmware and DDP package from
20  `https://www.intel.com/content/www/us/en/search.html?ws=text#q=e810&t=Downloads&layout=table`.
21
22- To understand what is DDP package and how it works, please review `Intel® Ethernet Controller E810 Dynamic
23  Device Personalization (DDP) for Telecommunications Technology Guide <https://cdrdv2.intel.com/v1/dl/getContent/617015>`_.
24
25- To understand DDP for COMMs usage with DPDK, please review `Intel® Ethernet 800 Series Telecommunication (Comms)
26  Dynamic Device Personalization (DDP) Package <https://cdrdv2.intel.com/v1/dl/getContent/618651>`_.
27
28Windows Prerequisites
29---------------------
30
31- Follow the DPDK `Getting Started Guide for Windows <https://doc.dpdk.org/guides/windows_gsg/index.html>`_ to setup the basic DPDK environment.
32
33- Identify the Intel® Ethernet adapter and get the latest NVM/FW version.
34
35- To access any Intel® Ethernet hardware, load the NetUIO driver in place of existing built-in (inbox) driver.
36
37- To load NetUIO driver, follow the steps mentioned in `dpdk-kmods repository
38  <https://git.dpdk.org/dpdk-kmods/tree/windows/netuio/README.rst>`_.
39
40- Loading of private Dynamic Device Personalization (DDP) package is not supported on Windows.
41
42
43Recommended Matching List
44-------------------------
45
46It is highly recommended to upgrade the ice kernel driver, firmware and DDP package
47to avoid the compatibility issues with ice PMD.
48Here is the suggested matching list which has been tested and verified.
49The detailed information can refer to chapter Tested Platforms/Tested NICs in release notes.
50
51   +-----------+---------------+-----------------+-----------+--------------+-----------+
52   |    DPDK   | Kernel Driver | OS Default DDP  | COMMS DDP | Wireless DDP | Firmware  |
53   +===========+===============+=================+===========+==============+===========+
54   |    20.11  |     1.3.2     |      1.3.20     |  1.3.24   |      N/A     |    2.3    |
55   +-----------+---------------+-----------------+-----------+--------------+-----------+
56   |    21.02  |     1.4.11    |      1.3.24     |  1.3.28   |    1.3.4     |    2.4    |
57   +-----------+---------------+-----------------+-----------+--------------+-----------+
58   |    21.05  |     1.6.5     |      1.3.26     |  1.3.30   |    1.3.6     |    3.0    |
59   +-----------+---------------+-----------------+-----------+--------------+-----------+
60
61Pre-Installation Configuration
62------------------------------
63
64
65Runtime Config Options
66~~~~~~~~~~~~~~~~~~~~~~
67
68- ``Safe Mode Support`` (default ``0``)
69
70  If driver failed to load OS package, by default driver's initialization failed.
71  But if user intend to use the device without OS package, user can take ``devargs``
72  parameter ``safe-mode-support``, for example::
73
74    -a 80:00.0,safe-mode-support=1
75
76  Then the driver will be initialized successfully and the device will enter Safe Mode.
77  NOTE: In Safe mode, only very limited features are available, features like RSS,
78  checksum, fdir, tunneling ... are all disabled.
79
80- ``Generic Flow Pipeline Mode Support`` (default ``0``)
81
82  In pipeline mode, a flow can be set at one specific stage by setting parameter
83  ``priority``. Currently, we support two stages: priority = 0 or !0. Flows with
84  priority 0 located at the first pipeline stage which typically be used as a firewall
85  to drop the packet on a blocklist(we called it permission stage). At this stage,
86  flow rules are created for the device's exact match engine: switch. Flows with priority
87  !0 located at the second stage, typically packets are classified here and be steered to
88  specific queue or queue group (we called it distribution stage), At this stage, flow
89  rules are created for device's flow director engine.
90  For none-pipeline mode, ``priority`` is ignored, a flow rule can be created as a flow director
91  rule or a switch rule depends on its pattern/action and the resource allocation situation,
92  all flows are virtually at the same pipeline stage.
93  By default, generic flow API is enabled in none-pipeline mode, user can choose to
94  use pipeline mode by setting ``devargs`` parameter ``pipeline-mode-support``,
95  for example::
96
97    -a 80:00.0,pipeline-mode-support=1
98
99- ``Protocol extraction for per queue``
100
101  Configure the RX queues to do protocol extraction into mbuf for protocol
102  handling acceleration, like checking the TCP SYN packets quickly.
103
104  The argument format is::
105
106      -a 18:00.0,proto_xtr=<queues:protocol>[<queues:protocol>...]
107      -a 18:00.0,proto_xtr=<protocol>
108
109  Queues are grouped by ``(`` and ``)`` within the group. The ``-`` character
110  is used as a range separator and ``,`` is used as a single number separator.
111  The grouping ``()`` can be omitted for single element group. If no queues are
112  specified, PMD will use this protocol extraction type for all queues.
113
114  Protocol is : ``vlan, ipv4, ipv6, ipv6_flow, tcp, ip_offset``.
115
116  .. code-block:: console
117
118    dpdk-testpmd -a 18:00.0,proto_xtr='[(1,2-3,8-9):tcp,10-13:vlan]'
119
120  This setting means queues 1, 2-3, 8-9 are TCP extraction, queues 10-13 are
121  VLAN extraction, other queues run with no protocol extraction.
122
123  .. code-block:: console
124
125    dpdk-testpmd -a 18:00.0,proto_xtr=vlan,proto_xtr='[(1,2-3,8-9):tcp,10-23:ipv6]'
126
127  This setting means queues 1, 2-3, 8-9 are TCP extraction, queues 10-23 are
128  IPv6 extraction, other queues use the default VLAN extraction.
129
130  The extraction metadata is copied into the registered dynamic mbuf field, and
131  the related dynamic mbuf flags is set.
132
133  .. table:: Protocol extraction : ``vlan``
134
135   +----------------------------+----------------------------+
136   |           VLAN2            |           VLAN1            |
137   +======+===+=================+======+===+=================+
138   |  PCP | D |       VID       |  PCP | D |       VID       |
139   +------+---+-----------------+------+---+-----------------+
140
141  VLAN1 - single or EVLAN (first for QinQ).
142
143  VLAN2 - C-VLAN (second for QinQ).
144
145  .. table:: Protocol extraction : ``ipv4``
146
147   +----------------------------+----------------------------+
148   |           IPHDR2           |           IPHDR1           |
149   +======+=======+=============+==============+=============+
150   |  Ver |Hdr Len|    ToS      |      TTL     |  Protocol   |
151   +------+-------+-------------+--------------+-------------+
152
153  IPHDR1 - IPv4 header word 4, "TTL" and "Protocol" fields.
154
155  IPHDR2 - IPv4 header word 0, "Ver", "Hdr Len" and "Type of Service" fields.
156
157  .. table:: Protocol extraction : ``ipv6``
158
159   +----------------------------+----------------------------+
160   |           IPHDR2           |           IPHDR1           |
161   +=====+=============+========+=============+==============+
162   | Ver |Traffic class|  Flow  | Next Header |   Hop Limit  |
163   +-----+-------------+--------+-------------+--------------+
164
165  IPHDR1 - IPv6 header word 3, "Next Header" and "Hop Limit" fields.
166
167  IPHDR2 - IPv6 header word 0, "Ver", "Traffic class" and high 4 bits of
168  "Flow Label" fields.
169
170  .. table:: Protocol extraction : ``ipv6_flow``
171
172   +----------------------------+----------------------------+
173   |           IPHDR2           |           IPHDR1           |
174   +=====+=============+========+============================+
175   | Ver |Traffic class|            Flow Label               |
176   +-----+-------------+-------------------------------------+
177
178  IPHDR1 - IPv6 header word 1, 16 low bits of the "Flow Label" field.
179
180  IPHDR2 - IPv6 header word 0, "Ver", "Traffic class" and high 4 bits of
181  "Flow Label" fields.
182
183  .. table:: Protocol extraction : ``tcp``
184
185   +----------------------------+----------------------------+
186   |           TCPHDR2          |           TCPHDR1          |
187   +============================+======+======+==============+
188   |          Reserved          |Offset|  RSV |     Flags    |
189   +----------------------------+------+------+--------------+
190
191  TCPHDR1 - TCP header word 6, "Data Offset" and "Flags" fields.
192
193  TCPHDR2 - Reserved
194
195  .. table:: Protocol extraction : ``ip_offset``
196
197   +----------------------------+----------------------------+
198   |           IPHDR2           |           IPHDR1           |
199   +============================+============================+
200   |       IPv6 HDR Offset      |       IPv4 HDR Offset      |
201   +----------------------------+----------------------------+
202
203  IPHDR1 - Outer/Single IPv4 Header offset.
204
205  IPHDR2 - Outer/Single IPv6 Header offset.
206
207  Use ``rte_net_ice_dynf_proto_xtr_metadata_get`` to access the protocol
208  extraction metadata, and use ``RTE_PKT_RX_DYNF_PROTO_XTR_*`` to get the
209  metadata type of ``struct rte_mbuf::ol_flags``.
210
211  The ``rte_net_ice_dump_proto_xtr_metadata`` routine shows how to
212  access the protocol extraction result in ``struct rte_mbuf``.
213
214- ``Hardware debug mask log support`` (default ``0``)
215
216  User can enable the related hardware debug mask such as ICE_DBG_NVM::
217
218    -a 0000:88:00.0,hw_debug_mask=0x80 --log-level=pmd.net.ice.driver:8
219
220  These ICE_DBG_XXX are defined in ``drivers/net/ice/base/ice_type.h``.
221
222Driver compilation and testing
223------------------------------
224
225Refer to the document :ref:`compiling and testing a PMD for a NIC <pmd_build_and_test>`
226for details.
227
228Features
229--------
230
231Vector PMD
232~~~~~~~~~~
233
234Vector PMD for RX and TX path are selected automatically. The paths
235are chosen based on 2 conditions.
236
237- ``CPU``
238  On the X86 platform, the driver checks if the CPU supports AVX2.
239  If it's supported, AVX2 paths will be chosen. If not, SSE is chosen.
240  If the CPU supports AVX512 and EAL argument ``--force-max-simd-bitwidth``
241  is set to 512, AVX512 paths will be chosen.
242
243- ``Offload features``
244  The supported HW offload features are described in the document ice.ini,
245  A value "P" means the offload feature is not supported by vector path.
246  If any not supported features are used, ICE vector PMD is disabled and the
247  normal paths are chosen.
248
249Malicious driver detection (MDD)
250~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
251
252It's not appropriate to send a packet, if this packet's destination MAC address
253is just this port's MAC address. If SW tries to send such packets, HW will
254report a MDD event and drop the packets.
255
256The APPs based on DPDK should avoid providing such packets.
257
258Device Config Function (DCF)
259~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
260
261This section demonstrates ICE DCF PMD, which shares the core module with ICE
262PMD and iAVF PMD.
263
264A DCF (Device Config Function) PMD bounds to the device's trusted VF with ID 0,
265it can act as a sole controlling entity to exercise advance functionality (such
266as switch, ACL) for the rest VFs.
267
268The DCF PMD needs to advertise and acquire DCF capability which allows DCF to
269send AdminQ commands that it would like to execute over to the PF and receive
270responses for the same from PF.
271
272.. _figure_ice_dcf:
273
274.. figure:: img/ice_dcf.*
275
276   DCF Communication flow.
277
278#. Create the VFs::
279
280      echo 4 > /sys/bus/pci/devices/0000\:18\:00.0/sriov_numvfs
281
282#. Enable the VF0 trust on::
283
284      ip link set dev enp24s0f0 vf 0 trust on
285
286#. Bind the VF0,  and run testpmd with 'cap=dcf' devarg::
287
288      dpdk-testpmd -l 22-25 -n 4 -a 18:01.0,cap=dcf -- -i
289
290#. Monitor the VF2 interface network traffic::
291
292      tcpdump -e -nn -i enp24s1f2
293
294#. Create one flow to redirect the traffic to VF2 by DCF::
295
296      flow create 0 priority 0 ingress pattern eth / ipv4 src is 192.168.0.2 \
297      dst is 192.168.0.3 / end actions vf id 2 / end
298
299#. Send the packet, and it should be displayed on tcpdump::
300
301      sendp(Ether(src='3c:fd:fe:aa:bb:78', dst='00:00:00:01:02:03')/IP(src=' \
302      192.168.0.2', dst="192.168.0.3")/TCP(flags='S')/Raw(load='XXXXXXXXXX'), \
303      iface="enp24s0f0", count=10)
304
305Sample Application Notes
306------------------------
307
308Vlan filter
309~~~~~~~~~~~
310
311Vlan filter only works when Promiscuous mode is off.
312
313To start ``testpmd``, and add vlan 10 to port 0:
314
315.. code-block:: console
316
317    ./app/dpdk-testpmd -l 0-15 -n 4 -- -i
318    ...
319
320    testpmd> rx_vlan add 10 0
321
322Limitations or Known issues
323---------------------------
324
325The Intel E810 requires a programmable pipeline package be downloaded
326by the driver to support normal operations. The E810 has a limited
327functionality built in to allow PXE boot and other use cases, but the
328driver must download a package file during the driver initialization
329stage.
330
331The default DDP package file name is ice.pkg. For a specific NIC, the
332DDP package supposed to be loaded can have a filename: ice-xxxxxx.pkg,
333where 'xxxxxx' is the 64-bit PCIe Device Serial Number of the NIC. For
334example, if the NIC's device serial number is 00-CC-BB-FF-FF-AA-05-68,
335the device-specific DDP package filename is ice-00ccbbffffaa0568.pkg
336(in hex and all low case). During initialization, the driver searches
337in the following paths in order: /lib/firmware/updates/intel/ice/ddp
338and /lib/firmware/intel/ice/ddp. The corresponding device-specific DDP
339package will be downloaded first if the file exists. If not, then the
340driver tries to load the default package. The type of loaded package
341is stored in ``ice_adapter->active_pkg_type``.
342
343A symbolic link to the DDP package file is also ok. The same package
344file is used by both the kernel driver and the DPDK PMD.
345
346   .. Note::
347
348      Windows support: The DDP package is not supported on Windows so,
349      loading of the package is disabled on Windows.
350