1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2015-2019 Intel Corporation.
4 Intel(R) QuickAssist (QAT) Crypto Poll Mode Driver
5 ==================================================
7 QAT documentation consists of three parts:
9 * Details of the symmetric crypto service below.
10 * Details of the `compression service <http://doc.dpdk.org/guides/compressdevs/qat_comp.html>`_
11 in the compressdev drivers section.
12 * Details of building the common QAT infrastructure and the PMDs to support the
13 above services. See :ref:`building_qat` below.
16 Symmetric Crypto Service on QAT
17 -------------------------------
19 The QAT crypto PMD provides poll mode crypto driver support for the following
20 hardware accelerator devices:
22 * ``Intel QuickAssist Technology DH895xCC``
23 * ``Intel QuickAssist Technology C62x``
24 * ``Intel QuickAssist Technology C3xxx``
25 * ``Intel QuickAssist Technology D15xx``
26 * ``Intel QuickAssist Technology C4xxx``
32 The QAT PMD has support for:
36 * ``RTE_CRYPTO_CIPHER_3DES_CBC``
37 * ``RTE_CRYPTO_CIPHER_3DES_CTR``
38 * ``RTE_CRYPTO_CIPHER_AES128_CBC``
39 * ``RTE_CRYPTO_CIPHER_AES192_CBC``
40 * ``RTE_CRYPTO_CIPHER_AES256_CBC``
41 * ``RTE_CRYPTO_CIPHER_AES128_CTR``
42 * ``RTE_CRYPTO_CIPHER_AES192_CTR``
43 * ``RTE_CRYPTO_CIPHER_AES256_CTR``
44 * ``RTE_CRYPTO_CIPHER_AES_XTS``
45 * ``RTE_CRYPTO_CIPHER_SNOW3G_UEA2``
46 * ``RTE_CRYPTO_CIPHER_NULL``
47 * ``RTE_CRYPTO_CIPHER_KASUMI_F8``
48 * ``RTE_CRYPTO_CIPHER_DES_CBC``
49 * ``RTE_CRYPTO_CIPHER_AES_DOCSISBPI``
50 * ``RTE_CRYPTO_CIPHER_DES_DOCSISBPI``
51 * ``RTE_CRYPTO_CIPHER_ZUC_EEA3``
55 * ``RTE_CRYPTO_AUTH_SHA1_HMAC``
56 * ``RTE_CRYPTO_AUTH_SHA224_HMAC``
57 * ``RTE_CRYPTO_AUTH_SHA256_HMAC``
58 * ``RTE_CRYPTO_AUTH_SHA384_HMAC``
59 * ``RTE_CRYPTO_AUTH_SHA512_HMAC``
60 * ``RTE_CRYPTO_AUTH_AES_XCBC_MAC``
61 * ``RTE_CRYPTO_AUTH_SNOW3G_UIA2``
62 * ``RTE_CRYPTO_AUTH_MD5_HMAC``
63 * ``RTE_CRYPTO_AUTH_NULL``
64 * ``RTE_CRYPTO_AUTH_KASUMI_F9``
65 * ``RTE_CRYPTO_AUTH_AES_GMAC``
66 * ``RTE_CRYPTO_AUTH_ZUC_EIA3``
67 * ``RTE_CRYPTO_AUTH_AES_CMAC``
69 Supported AEAD algorithms:
71 * ``RTE_CRYPTO_AEAD_AES_GCM``
72 * ``RTE_CRYPTO_AEAD_AES_CCM``
78 * Only supports the session-oriented API implementation (session-less APIs are not supported).
79 * SNOW 3G (UEA2), KASUMI (F8) and ZUC (EEA3) supported only if cipher length and offset fields are byte-multiple.
80 * SNOW 3G (UIA2) and ZUC (EIA3) supported only if hash length and offset fields are byte-multiple.
81 * No BSD support as BSD QAT kernel driver not available.
82 * ZUC EEA3/EIA3 is not supported by dh895xcc devices
83 * Maximum additional authenticated data (AAD) for GCM is 240 bytes long and must be passed to the device in a buffer rounded up to the nearest block-size multiple (x16) and padded with zeros.
84 * Queue pairs are not thread-safe (that is, within a single queue pair, RX and TX from different lcores is not supported).
86 Extra notes on KASUMI F9
87 ~~~~~~~~~~~~~~~~~~~~~~~~
89 When using KASUMI F9 authentication algorithm, the input buffer must be
90 constructed according to the
91 `3GPP KASUMI specification <http://cryptome.org/3gpp/35201-900.pdf>`_
92 (section 4.4, page 13). The input buffer has to have COUNT (4 bytes),
93 FRESH (4 bytes), MESSAGE and DIRECTION (1 bit) concatenated. After the DIRECTION
94 bit, a single '1' bit is appended, followed by between 0 and 7 '0' bits, so that
95 the total length of the buffer is multiple of 8 bits. Note that the actual
96 message can be any length, specified in bits.
98 Once this buffer is passed this way, when creating the crypto operation,
99 length of data to authenticate "op.sym.auth.data.length" must be the length
100 of all the items described above, including the padding at the end.
101 Also, offset of data to authenticate "op.sym.auth.data.offset"
102 must be such that points at the start of the COUNT bytes.
104 Asymmetric Crypto Service on QAT
105 --------------------------------
107 The QAT Asym PMD has support for:
109 * ``Modular exponentiation``
110 * ``Modular multiplicative inverse``
120 A QAT device can host multiple acceleration services:
122 * symmetric cryptography
124 * asymmetric cryptography
126 These services are provided to DPDK applications via PMDs which register to
127 implement the corresponding cryptodev and compressdev APIs. The PMDs use
128 common QAT driver code which manages the QAT PCI device. They also depend on a
129 QAT kernel driver being installed on the platform, see :ref:`qat_kernel` below.
132 Configuring and Building the DPDK QAT PMDs
133 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
136 Further information on configuring, building and installing DPDK is described
137 `here <http://doc.dpdk.org/guides/linux_gsg/build_dpdk.html>`_.
140 Quick instructions for QAT cryptodev PMD are as follows:
142 .. code-block:: console
144 cd to the top-level DPDK directory
146 sed -i 's,\(CONFIG_RTE_LIBRTE_PMD_QAT_SYM\)=n,\1=y,' build/.config
149 Quick instructions for QAT compressdev PMD are as follows:
151 .. code-block:: console
153 cd to the top-level DPDK directory
158 .. _building_qat_config:
163 These are the build configuration options affecting QAT, and their default values:
165 .. code-block:: console
167 CONFIG_RTE_LIBRTE_PMD_QAT=y
168 CONFIG_RTE_LIBRTE_PMD_QAT_SYM=n
169 CONFIG_RTE_PMD_QAT_MAX_PCI_DEVICES=48
170 CONFIG_RTE_PMD_QAT_COMP_IM_BUFFER_SIZE=65536
172 CONFIG_RTE_LIBRTE_PMD_QAT must be enabled for any QAT PMD to be built.
174 The QAT cryptodev PMD has an external dependency on libcrypto, so is not
175 built by default. CONFIG_RTE_LIBRTE_PMD_QAT_SYM should be enabled to build it.
177 The QAT compressdev PMD has no external dependencies, so needs no configuration
178 options and is built by default.
180 The number of VFs per PF varies - see table below. If multiple QAT packages are
181 installed on a platform then CONFIG_RTE_PMD_QAT_MAX_PCI_DEVICES should be
182 adjusted to the number of VFs which the QAT common code will need to handle.
183 Note, there are separate config items for max cryptodevs CONFIG_RTE_CRYPTO_MAX_DEVS
184 and max compressdevs CONFIG_RTE_COMPRESS_MAX_DEVS, if necessary these should be
185 adjusted to handle the total of QAT and other devices which the process will use.
187 QAT compression PMD needs intermediate buffers to support Deflate compression
188 with Dynamic Huffman encoding. CONFIG_RTE_PMD_QAT_COMP_IM_BUFFER_SIZE
189 specifies the size of a single buffer, the PMD will allocate a multiple of these,
190 plus some extra space for associated meta-data. For GEN2 devices, 20 buffers are
191 allocated while for GEN1 devices, 12 buffers are allocated, plus 1472 bytes overhead.
195 If the compressed output of a Deflate operation using Dynamic Huffman
196 Encoding is too big to fit in an intermediate buffer, then the
197 operation will fall back to fixed compression rather than failing the operation.
198 To avoid this less performant case, applications should configure
199 the intermediate buffer size to be larger than the expected input data size
200 (compressed output size is usually unknown, so the only option is to make
201 larger than the input size).
204 Device and driver naming
205 ~~~~~~~~~~~~~~~~~~~~~~~~
207 * The qat cryptodev driver name is "crypto_qat".
208 The "rte_cryptodev_devices_get()" returns the devices exposed by this driver.
210 * Each qat crypto device has a unique name, in format
211 "<pci bdf>_<service>", e.g. "0000:41:01.0_qat_sym".
212 This name can be passed to "rte_cryptodev_get_dev_id()" to get the device_id.
216 The qat crypto driver name is passed to the dpdk-test-crypto-perf tool in the "-devtype" parameter.
218 The qat crypto device name is in the format of the slave parameter passed to the crypto scheduler.
220 * The qat compressdev driver name is "compress_qat".
221 The rte_compressdev_devices_get() returns the devices exposed by this driver.
223 * Each qat compression device has a unique name, in format
224 <pci bdf>_<service>, e.g. "0000:41:01.0_qat_comp".
225 This name can be passed to rte_compressdev_get_dev_id() to get the device_id.
229 Dependency on the QAT kernel driver
230 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
232 To use QAT an SRIOV-enabled QAT kernel driver is required. The VF
233 devices created and initialised by this driver will be used by the QAT PMDs.
235 Instructions for installation are below, but first an explanation of the
236 relationships between the PF/VF devices and the PMDs visible to
239 Each QuickAssist PF device exposes a number of VF devices. Each VF device can
240 enable one cryptodev PMD and/or one compressdev PMD.
241 These QAT PMDs share the same underlying device and pci-mgmt code, but are
242 enumerated independently on their respective APIs and appear as independent
243 devices to applications.
247 Each VF can only be used by one DPDK process. It is not possible to share
248 the same VF across multiple processes, even if these processes are using
249 different acceleration services.
251 Conversely one DPDK process can use one or more QAT VFs and can expose both
252 cryptodev and compressdev instances on each of those VFs.
255 Available kernel drivers
256 ~~~~~~~~~~~~~~~~~~~~~~~~
258 Kernel drivers for each device for each service are listed in the following table. (Scroll right
259 to see the full table)
262 .. _table_qat_pmds_drivers:
264 .. table:: QAT device generations, devices and drivers
266 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
267 | S | A | C | Gen | Device | Driver/ver | Kernel Module | Pci Driver | PF Did | #PFs | VF Did | VFs/PF |
268 +=====+=====+=====+=====+==========+===============+===============+============+========+======+========+========+
269 | Yes | No | No | 1 | DH895xCC | linux/4.4+ | qat_dh895xcc | dh895xcc | 435 | 1 | 443 | 32 |
270 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
271 | Yes | No | No | " | " | 01.org/4.2.0+ | " | " | " | " | " | " |
272 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
273 | Yes | No | Yes | " | " | 01.org/4.3.0+ | " | " | " | " | " | " |
274 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
275 | Yes | No | No | 2 | C62x | linux/4.5+ | qat_c62x | c6xx | 37c8 | 3 | 37c9 | 16 |
276 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
277 | Yes | No | Yes | " | " | 01.org/4.2.0+ | " | " | " | " | " | " |
278 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
279 | Yes | No | No | 2 | C3xxx | linux/4.5+ | qat_c3xxx | c3xxx | 19e2 | 1 | 19e3 | 16 |
280 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
281 | Yes | No | Yes | " | " | 01.org/4.2.0+ | " | " | " | " | " | " |
282 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
283 | Yes | No | No | 2 | D15xx | p | qat_d15xx | d15xx | 6f54 | 1 | 6f55 | 16 |
284 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
285 | Yes | No | No | 3 | C4xxx | p | qat_c4xxx | c4xxx | 18a0 | 1 | 18a1 | 128 |
286 +-----+-----+-----+-----+----------+---------------+---------------+------------+--------+------+--------+--------+
288 The first 3 columns indicate the service:
290 * S = Symmetric crypto service (via cryptodev API)
291 * A = Asymmetric crypto service (via cryptodev API)
292 * C = Compression service (via compressdev API)
294 The ``Driver`` column indicates either the Linux kernel version in which
295 support for this device was introduced or a driver available on Intel's 01.org
296 website. There are both linux in-tree and 01.org kernel drivers available for some
297 devices. p = release pending.
299 If you are running on a kernel which includes a driver for your device, see
300 `Installation using kernel.org driver`_ below. Otherwise see
301 `Installation using 01.org QAT driver`_.
304 Installation using kernel.org driver
305 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
307 The examples below are based on the C62x device, if you have a different device
308 use the corresponding values in the above table.
310 In BIOS ensure that SRIOV is enabled and either:
313 * Enable VT-d and set ``"intel_iommu=on iommu=pt"`` in the grub file.
315 Check that the QAT driver is loaded on your system, by executing::
319 You should see the kernel module for your device listed, e.g.::
322 intel_qat 82336 1 qat_c62x
324 Next, you need to expose the Virtual Functions (VFs) using the sysfs file system.
326 First find the BDFs (Bus-Device-Function) of the physical functions (PFs) of
331 You should see output similar to::
333 1a:00.0 Co-processor: Intel Corporation Device 37c8
334 3d:00.0 Co-processor: Intel Corporation Device 37c8
335 3f:00.0 Co-processor: Intel Corporation Device 37c8
337 Enable the VFs for each PF by echoing the number of VFs per PF to the pci driver::
339 echo 16 > /sys/bus/pci/drivers/c6xx/0000:1a:00.0/sriov_numvfs
340 echo 16 > /sys/bus/pci/drivers/c6xx/0000:3d:00.0/sriov_numvfs
341 echo 16 > /sys/bus/pci/drivers/c6xx/0000:3f:00.0/sriov_numvfs
343 Check that the VFs are available for use. For example ``lspci -d:37c9`` should
344 list 48 VF devices available for a ``C62x`` device.
346 To complete the installation follow the instructions in
347 `Binding the available VFs to the DPDK UIO driver`_.
351 If the QAT kernel modules are not loaded and you see an error like ``Failed
352 to load MMP firmware qat_895xcc_mmp.bin`` in kernel logs, this may be as a
353 result of not using a distribution, but just updating the kernel directly.
355 Download firmware from the `kernel firmware repo
356 <http://git.kernel.org/cgit/linux/kernel/git/firmware/linux-firmware.git/tree/>`_.
358 Copy qat binaries to ``/lib/firmware``::
360 cp qat_895xcc.bin /lib/firmware
361 cp qat_895xcc_mmp.bin /lib/firmware
363 Change to your linux source root directory and start the qat kernel modules::
365 insmod ./drivers/crypto/qat/qat_common/intel_qat.ko
366 insmod ./drivers/crypto/qat/qat_dh895xcc/qat_dh895xcc.ko
371 If you see the following warning in ``/var/log/messages`` it can be ignored:
372 ``IOMMU should be enabled for SR-IOV to work correctly``.
375 Installation using 01.org QAT driver
376 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
378 Download the latest QuickAssist Technology Driver from `01.org
379 <https://01.org/packet-processing/intel%C2%AE-quickassist-technology-drivers-and-patches>`_.
380 Consult the *Getting Started Guide* at the same URL for further information.
382 The steps below assume you are:
384 * Building on a platform with one ``C62x`` device.
385 * Using package ``qat1.7.l.4.2.0-000xx.tar.gz``.
386 * On Fedora26 kernel ``4.11.11-300.fc26.x86_64``.
388 In the BIOS ensure that SRIOV is enabled and VT-d is disabled.
390 Uninstall any existing QAT driver, for example by running:
392 * ``./installer.sh uninstall`` in the directory where originally installed.
395 Build and install the SRIOV-enabled QAT driver::
400 # Copy the package to this location and unpack
401 tar zxof qat1.7.l.4.2.0-000xx.tar.gz
403 ./configure --enable-icp-sriov=host
406 You can use ``cat /sys/kernel/debug/qat<your device type and bdf>/version/fw`` to confirm the driver is correctly installed and is using firmware version 4.2.0.
407 You can use ``lspci -d:37c9`` to confirm the presence of the 16 VF devices available per ``C62x`` PF.
409 Confirm the driver is correctly installed and is using firmware version 4.2.0::
411 cat /sys/kernel/debug/qat<your device type and bdf>/version/fw
414 Confirm the presence of 48 VF devices - 16 per PF::
419 To complete the installation - follow instructions in `Binding the available VFs to the DPDK UIO driver`_.
423 If using a later kernel and the build fails with an error relating to
424 ``strict_stroul`` not being available apply the following patch:
428 /QAT/QAT1.6/quickassist/utilities/downloader/Target_CoreLibs/uclo/include/linux/uclo_platform.h
429 + #if LINUX_VERSION_CODE >= KERNEL_VERSION(3,18,5)
430 + #define STR_TO_64(str, base, num, endPtr) {endPtr=NULL; if (kstrtoul((str), (base), (num))) printk("Error strtoull convert %s\n", str); }
432 #if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,38)
433 #define STR_TO_64(str, base, num, endPtr) {endPtr=NULL; if (strict_strtoull((str), (base), (num))) printk("Error strtoull convert %s\n", str); }
435 #if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
436 #define STR_TO_64(str, base, num, endPtr) {endPtr=NULL; strict_strtoll((str), (base), (num));}
438 #define STR_TO_64(str, base, num, endPtr) \
442 *(num) = -(simple_strtoull((str+1), &(endPtr), (base))); \
444 *(num) = simple_strtoull((str), &(endPtr), (base)); \
454 If the build fails due to missing header files you may need to do following::
456 sudo yum install zlib-devel
457 sudo yum install openssl-devel
458 sudo yum install libudev-devel
462 If the build or install fails due to mismatching kernel sources you may need to do the following::
464 sudo yum install kernel-headers-`uname -r`
465 sudo yum install kernel-src-`uname -r`
466 sudo yum install kernel-devel-`uname -r`
469 Binding the available VFs to the DPDK UIO driver
470 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
472 Unbind the VFs from the stock driver so they can be bound to the uio driver.
474 For an Intel(R) QuickAssist Technology DH895xCC device
475 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
477 The unbind command below assumes ``BDFs`` of ``03:01.00-03:04.07``, if your
478 VFs are different adjust the unbind command below::
480 for device in $(seq 1 4); do \
481 for fn in $(seq 0 7); do \
482 echo -n 0000:03:0${device}.${fn} > \
483 /sys/bus/pci/devices/0000\:03\:0${device}.${fn}/driver/unbind; \
487 For an Intel(R) QuickAssist Technology C62x device
488 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
490 The unbind command below assumes ``BDFs`` of ``1a:01.00-1a:02.07``,
491 ``3d:01.00-3d:02.07`` and ``3f:01.00-3f:02.07``, if your VFs are different
492 adjust the unbind command below::
494 for device in $(seq 1 2); do \
495 for fn in $(seq 0 7); do \
496 echo -n 0000:1a:0${device}.${fn} > \
497 /sys/bus/pci/devices/0000\:1a\:0${device}.${fn}/driver/unbind; \
499 echo -n 0000:3d:0${device}.${fn} > \
500 /sys/bus/pci/devices/0000\:3d\:0${device}.${fn}/driver/unbind; \
502 echo -n 0000:3f:0${device}.${fn} > \
503 /sys/bus/pci/devices/0000\:3f\:0${device}.${fn}/driver/unbind; \
507 For Intel(R) QuickAssist Technology C3xxx or D15xx device
508 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
510 The unbind command below assumes ``BDFs`` of ``01:01.00-01:02.07``, if your
511 VFs are different adjust the unbind command below::
513 for device in $(seq 1 2); do \
514 for fn in $(seq 0 7); do \
515 echo -n 0000:01:0${device}.${fn} > \
516 /sys/bus/pci/devices/0000\:01\:0${device}.${fn}/driver/unbind; \
520 Bind to the DPDK uio driver
521 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
523 Install the DPDK igb_uio driver, bind the VF PCI Device id to it and use lspci
524 to confirm the VF devices are now in use by igb_uio kernel driver,
525 e.g. for the C62x device::
527 cd to the top-level DPDK directory
529 insmod ./build/kmod/igb_uio.ko
530 echo "8086 37c9" > /sys/bus/pci/drivers/igb_uio/new_id
534 Another way to bind the VFs to the DPDK UIO driver is by using the
535 ``dpdk-devbind.py`` script::
537 cd to the top-level DPDK directory
538 ./usertools/dpdk-devbind.py -b igb_uio 0000:03:01.1
543 QAT crypto PMD can be tested by running the test application::
548 ./test -l1 -n1 -w <your qat bdf>
549 RTE>>cryptodev_qat_autotest
551 QAT compression PMD can be tested by running the test application::
554 sed -i 's,\(CONFIG_RTE_COMPRESSDEV_TEST\)=n,\1=y,' build/.config
557 ./test -l1 -n1 -w <your qat bdf>
558 RTE>>compressdev_autotest
564 There are 2 sets of trace available via the dynamic logging feature:
566 * pmd.qat_dp exposes trace on the data-path.
567 * pmd.qat_general exposes all other trace.
569 pmd.qat exposes both sets of traces.
570 They can be enabled using the log-level option (where 8=maximum log level) on
571 the process cmdline, e.g. using any of the following::
573 --log-level="pmd.qat_general,8"
574 --log-level="pmd.qat_dp,8"
575 --log-level="pmd.qat,8"
579 The global RTE_LOG_DP_LEVEL overrides data-path trace so must be set to
580 RTE_LOG_DEBUG to see all the trace. This variable is in config/rte_config.h
581 for meson build and config/common_base for gnu make.
582 Also the dynamic global log level overrides both sets of trace, so e.g. no
583 QAT trace would display in this case::
585 --log-level="7" --log-level="pmd.qat_general,8"