Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
tjh.dev
kernel
os
linux
1.. SPDX-License-Identifier: GPL-2.0
2
3===================
4ice devlink support
5===================
6
7This document describes the devlink features implemented by the ``ice``
8device driver.
9
10Info versions
11=============
12
13The ``ice`` driver reports the following versions
14
15.. list-table:: devlink info versions implemented
16 :widths: 5 5 5 90
17
18 * - Name
19 - Type
20 - Example
21 - Description
22 * - ``board.id``
23 - fixed
24 - K65390-000
25 - The Product Board Assembly (PBA) identifier of the board.
26 * - ``fw.mgmt``
27 - running
28 - 2.1.7
29 - 3-digit version number of the management firmware running on the
30 Embedded Management Processor of the device. It controls the PHY,
31 link, access to device resources, etc. Intel documentation refers to
32 this as the EMP firmware.
33 * - ``fw.mgmt.api``
34 - running
35 - 1.5.1
36 - 3-digit version number (major.minor.patch) of the API exported over
37 the AdminQ by the management firmware. Used by the driver to
38 identify what commands are supported. Historical versions of the
39 kernel only displayed a 2-digit version number (major.minor).
40 * - ``fw.mgmt.build``
41 - running
42 - 0x305d955f
43 - Unique identifier of the source for the management firmware.
44 * - ``fw.undi``
45 - running
46 - 1.2581.0
47 - Version of the Option ROM containing the UEFI driver. The version is
48 reported in ``major.minor.patch`` format. The major version is
49 incremented whenever a major breaking change occurs, or when the
50 minor version would overflow. The minor version is incremented for
51 non-breaking changes and reset to 1 when the major version is
52 incremented. The patch version is normally 0 but is incremented when
53 a fix is delivered as a patch against an older base Option ROM.
54 * - ``fw.psid.api``
55 - running
56 - 0.80
57 - Version defining the format of the flash contents.
58 * - ``fw.bundle_id``
59 - running
60 - 0x80002ec0
61 - Unique identifier of the firmware image file that was loaded onto
62 the device. Also referred to as the EETRACK identifier of the NVM.
63 * - ``fw.app.name``
64 - running
65 - ICE OS Default Package
66 - The name of the DDP package that is active in the device. The DDP
67 package is loaded by the driver during initialization. Each
68 variation of the DDP package has a unique name.
69 * - ``fw.app``
70 - running
71 - 1.3.1.0
72 - The version of the DDP package that is active in the device. Note
73 that both the name (as reported by ``fw.app.name``) and version are
74 required to uniquely identify the package.
75 * - ``fw.app.bundle_id``
76 - running
77 - 0xc0000001
78 - Unique identifier for the DDP package loaded in the device. Also
79 referred to as the DDP Track ID. Can be used to uniquely identify
80 the specific DDP package.
81 * - ``fw.netlist``
82 - running
83 - 1.1.2000-6.7.0
84 - The version of the netlist module. This module defines the device's
85 Ethernet capabilities and default settings, and is used by the
86 management firmware as part of managing link and device
87 connectivity.
88 * - ``fw.netlist.build``
89 - running
90 - 0xee16ced7
91 - The first 4 bytes of the hash of the netlist module contents.
92
93Flash Update
94============
95
96The ``ice`` driver implements support for flash update using the
97``devlink-flash`` interface. It supports updating the device flash using a
98combined flash image that contains the ``fw.mgmt``, ``fw.undi``, and
99``fw.netlist`` components.
100
101.. list-table:: List of supported overwrite modes
102 :widths: 5 95
103
104 * - Bits
105 - Behavior
106 * - ``DEVLINK_FLASH_OVERWRITE_SETTINGS``
107 - Do not preserve settings stored in the flash components being
108 updated. This includes overwriting the port configuration that
109 determines the number of physical functions the device will
110 initialize with.
111 * - ``DEVLINK_FLASH_OVERWRITE_SETTINGS`` and ``DEVLINK_FLASH_OVERWRITE_IDENTIFIERS``
112 - Do not preserve either settings or identifiers. Overwrite everything
113 in the flash with the contents from the provided image, without
114 performing any preservation. This includes overwriting device
115 identifying fields such as the MAC address, VPD area, and device
116 serial number. It is expected that this combination be used with an
117 image customized for the specific device.
118
119The ice hardware does not support overwriting only identifiers while
120preserving settings, and thus ``DEVLINK_FLASH_OVERWRITE_IDENTIFIERS`` on its
121own will be rejected. If no overwrite mask is provided, the firmware will be
122instructed to preserve all settings and identifying fields when updating.
123
124Reload
125======
126
127The ``ice`` driver supports activating new firmware after a flash update
128using ``DEVLINK_CMD_RELOAD`` with the ``DEVLINK_RELOAD_ACTION_FW_ACTIVATE``
129action.
130
131.. code:: shell
132
133 $ devlink dev reload pci/0000:01:00.0 reload action fw_activate
134
135The new firmware is activated by issuing a device specific Embedded
136Management Processor reset which requests the device to reset and reload the
137EMP firmware image.
138
139The driver does not currently support reloading the driver via
140``DEVLINK_RELOAD_ACTION_DRIVER_REINIT``.
141
142Port split
143==========
144
145The ``ice`` driver supports port splitting only for port 0, as the FW has
146a predefined set of available port split options for the whole device.
147
148A system reboot is required for port split to be applied.
149
150The following command will select the port split option with 4 ports:
151
152.. code:: shell
153
154 $ devlink port split pci/0000:16:00.0/0 count 4
155
156The list of all available port options will be printed to dynamic debug after
157each ``split`` and ``unsplit`` command. The first option is the default.
158
159.. code:: shell
160
161 ice 0000:16:00.0: Available port split options and max port speeds (Gbps):
162 ice 0000:16:00.0: Status Split Quad 0 Quad 1
163 ice 0000:16:00.0: count L0 L1 L2 L3 L4 L5 L6 L7
164 ice 0000:16:00.0: Active 2 100 - - - 100 - - -
165 ice 0000:16:00.0: 2 50 - 50 - - - - -
166 ice 0000:16:00.0: Pending 4 25 25 25 25 - - - -
167 ice 0000:16:00.0: 4 25 25 - - 25 25 - -
168 ice 0000:16:00.0: 8 10 10 10 10 10 10 10 10
169 ice 0000:16:00.0: 1 100 - - - - - - -
170
171There could be multiple FW port options with the same port split count. When
172the same port split count request is issued again, the next FW port option with
173the same port split count will be selected.
174
175``devlink port unsplit`` will select the option with a split count of 1. If
176there is no FW option available with split count 1, you will receive an error.
177
178Regions
179=======
180
181The ``ice`` driver implements the following regions for accessing internal
182device data.
183
184.. list-table:: regions implemented
185 :widths: 15 85
186
187 * - Name
188 - Description
189 * - ``nvm-flash``
190 - The contents of the entire flash chip, sometimes referred to as
191 the device's Non Volatile Memory.
192 * - ``shadow-ram``
193 - The contents of the Shadow RAM, which is loaded from the beginning
194 of the flash. Although the contents are primarily from the flash,
195 this area also contains data generated during device boot which is
196 not stored in flash.
197 * - ``device-caps``
198 - The contents of the device firmware's capabilities buffer. Useful to
199 determine the current state and configuration of the device.
200
201Both the ``nvm-flash`` and ``shadow-ram`` regions can be accessed without a
202snapshot. The ``device-caps`` region requires a snapshot as the contents are
203sent by firmware and can't be split into separate reads.
204
205Users can request an immediate capture of a snapshot for all three regions
206via the ``DEVLINK_CMD_REGION_NEW`` command.
207
208.. code:: shell
209
210 $ devlink region show
211 pci/0000:01:00.0/nvm-flash: size 10485760 snapshot [] max 1
212 pci/0000:01:00.0/device-caps: size 4096 snapshot [] max 10
213
214 $ devlink region new pci/0000:01:00.0/nvm-flash snapshot 1
215 $ devlink region dump pci/0000:01:00.0/nvm-flash snapshot 1
216
217 $ devlink region dump pci/0000:01:00.0/nvm-flash snapshot 1
218 0000000000000000 0014 95dc 0014 9514 0035 1670 0034 db30
219 0000000000000010 0000 0000 ffff ff04 0029 8c00 0028 8cc8
220 0000000000000020 0016 0bb8 0016 1720 0000 0000 c00f 3ffc
221 0000000000000030 bada cce5 bada cce5 bada cce5 bada cce5
222
223 $ devlink region read pci/0000:01:00.0/nvm-flash snapshot 1 address 0 length 16
224 0000000000000000 0014 95dc 0014 9514 0035 1670 0034 db30
225
226 $ devlink region delete pci/0000:01:00.0/nvm-flash snapshot 1
227
228 $ devlink region new pci/0000:01:00.0/device-caps snapshot 1
229 $ devlink region dump pci/0000:01:00.0/device-caps snapshot 1
230 0000000000000000 01 00 01 00 00 00 00 00 01 00 00 00 00 00 00 00
231 0000000000000010 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
232 0000000000000020 02 00 02 01 32 03 00 00 0a 00 00 00 25 00 00 00
233 0000000000000030 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
234 0000000000000040 04 00 01 00 01 00 00 00 00 00 00 00 00 00 00 00
235 0000000000000050 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
236 0000000000000060 05 00 01 00 03 00 00 00 00 00 00 00 00 00 00 00
237 0000000000000070 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
238 0000000000000080 06 00 01 00 01 00 00 00 00 00 00 00 00 00 00 00
239 0000000000000090 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
240 00000000000000a0 08 00 01 00 00 00 00 00 00 00 00 00 00 00 00 00
241 00000000000000b0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
242 00000000000000c0 12 00 01 00 01 00 00 00 01 00 01 00 00 00 00 00
243 00000000000000d0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
244 00000000000000e0 13 00 01 00 00 01 00 00 00 00 00 00 00 00 00 00
245 00000000000000f0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
246 0000000000000100 14 00 01 00 01 00 00 00 00 00 00 00 00 00 00 00
247 0000000000000110 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
248 0000000000000120 15 00 01 00 01 00 00 00 00 00 00 00 00 00 00 00
249 0000000000000130 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
250 0000000000000140 16 00 01 00 01 00 00 00 00 00 00 00 00 00 00 00
251 0000000000000150 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
252 0000000000000160 17 00 01 00 06 00 00 00 00 00 00 00 00 00 00 00
253 0000000000000170 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
254 0000000000000180 18 00 01 00 01 00 00 00 01 00 00 00 08 00 00 00
255 0000000000000190 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
256 00000000000001a0 22 00 01 00 01 00 00 00 00 00 00 00 00 00 00 00
257 00000000000001b0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
258 00000000000001c0 40 00 01 00 00 08 00 00 08 00 00 00 00 00 00 00
259 00000000000001d0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
260 00000000000001e0 41 00 01 00 00 08 00 00 00 00 00 00 00 00 00 00
261 00000000000001f0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
262 0000000000000200 42 00 01 00 00 08 00 00 00 00 00 00 00 00 00 00
263 0000000000000210 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
264
265 $ devlink region delete pci/0000:01:00.0/device-caps snapshot 1
266
267Devlink Rate
268============
269
270The ``ice`` driver implements devlink-rate API. It allows for offload of
271the Hierarchical QoS to the hardware. It enables user to group Virtual
272Functions in a tree structure and assign supported parameters: tx_share,
273tx_max, tx_priority and tx_weight to each node in a tree. So effectively
274user gains an ability to control how much bandwidth is allocated for each
275VF group. This is later enforced by the HW.
276
277It is assumed that this feature is mutually exclusive with DCB performed
278in FW and ADQ, or any driver feature that would trigger changes in QoS,
279for example creation of the new traffic class. The driver will prevent DCB
280or ADQ configuration if user started making any changes to the nodes using
281devlink-rate API. To configure those features a driver reload is necessary.
282Correspondingly if ADQ or DCB will get configured the driver won't export
283hierarchy at all, or will remove the untouched hierarchy if those
284features are enabled after the hierarchy is exported, but before any
285changes are made.
286
287This feature is also dependent on switchdev being enabled in the system.
288It's required bacause devlink-rate requires devlink-port objects to be
289present, and those objects are only created in switchdev mode.
290
291If the driver is set to the switchdev mode, it will export internal
292hierarchy the moment VF's are created. Root of the tree is always
293represented by the node_0. This node can't be deleted by the user. Leaf
294nodes and nodes with children also can't be deleted.
295
296.. list-table:: Attributes supported
297 :widths: 15 85
298
299 * - Name
300 - Description
301 * - ``tx_max``
302 - maximum bandwidth to be consumed by the tree Node. Rate Limit is
303 an absolute number specifying a maximum amount of bytes a Node may
304 consume during the course of one second. Rate limit guarantees
305 that a link will not oversaturate the receiver on the remote end
306 and also enforces an SLA between the subscriber and network
307 provider.
308 * - ``tx_share``
309 - minimum bandwidth allocated to a tree node when it is not blocked.
310 It specifies an absolute BW. While tx_max defines the maximum
311 bandwidth the node may consume, the tx_share marks committed BW
312 for the Node.
313 * - ``tx_priority``
314 - allows for usage of strict priority arbiter among siblings. This
315 arbitration scheme attempts to schedule nodes based on their
316 priority as long as the nodes remain within their bandwidth limit.
317 Range 0-7. Nodes with priority 7 have the highest priority and are
318 selected first, while nodes with priority 0 have the lowest
319 priority. Nodes that have the same priority are treated equally.
320 * - ``tx_weight``
321 - allows for usage of Weighted Fair Queuing arbitration scheme among
322 siblings. This arbitration scheme can be used simultaneously with
323 the strict priority. Range 1-200. Only relative values mater for
324 arbitration.
325
326``tx_priority`` and ``tx_weight`` can be used simultaneously. In that case
327nodes with the same priority form a WFQ subgroup in the sibling group
328and arbitration among them is based on assigned weights.
329
330.. code:: shell
331
332 # enable switchdev
333 $ devlink dev eswitch set pci/0000:4b:00.0 mode switchdev
334
335 # at this point driver should export internal hierarchy
336 $ echo 2 > /sys/class/net/ens785np0/device/sriov_numvfs
337
338 $ devlink port function rate show
339 pci/0000:4b:00.0/node_25: type node parent node_24
340 pci/0000:4b:00.0/node_24: type node parent node_0
341 pci/0000:4b:00.0/node_32: type node parent node_31
342 pci/0000:4b:00.0/node_31: type node parent node_30
343 pci/0000:4b:00.0/node_30: type node parent node_16
344 pci/0000:4b:00.0/node_19: type node parent node_18
345 pci/0000:4b:00.0/node_18: type node parent node_17
346 pci/0000:4b:00.0/node_17: type node parent node_16
347 pci/0000:4b:00.0/node_14: type node parent node_5
348 pci/0000:4b:00.0/node_5: type node parent node_3
349 pci/0000:4b:00.0/node_13: type node parent node_4
350 pci/0000:4b:00.0/node_12: type node parent node_4
351 pci/0000:4b:00.0/node_11: type node parent node_4
352 pci/0000:4b:00.0/node_10: type node parent node_4
353 pci/0000:4b:00.0/node_9: type node parent node_4
354 pci/0000:4b:00.0/node_8: type node parent node_4
355 pci/0000:4b:00.0/node_7: type node parent node_4
356 pci/0000:4b:00.0/node_6: type node parent node_4
357 pci/0000:4b:00.0/node_4: type node parent node_3
358 pci/0000:4b:00.0/node_3: type node parent node_16
359 pci/0000:4b:00.0/node_16: type node parent node_15
360 pci/0000:4b:00.0/node_15: type node parent node_0
361 pci/0000:4b:00.0/node_2: type node parent node_1
362 pci/0000:4b:00.0/node_1: type node parent node_0
363 pci/0000:4b:00.0/node_0: type node
364 pci/0000:4b:00.0/1: type leaf parent node_25
365 pci/0000:4b:00.0/2: type leaf parent node_25
366
367 # let's create some custom node
368 $ devlink port function rate add pci/0000:4b:00.0/node_custom parent node_0
369
370 # second custom node
371 $ devlink port function rate add pci/0000:4b:00.0/node_custom_1 parent node_custom
372
373 # reassign second VF to newly created branch
374 $ devlink port function rate set pci/0000:4b:00.0/2 parent node_custom_1
375
376 # assign tx_weight to the VF
377 $ devlink port function rate set pci/0000:4b:00.0/2 tx_weight 5
378
379 # assign tx_share to the VF
380 $ devlink port function rate set pci/0000:4b:00.0/2 tx_share 500Mbps