```
===============================================================================
November 2022
===============================================================================
Quick References
================
- https://intel.com/quickassist (overview)
https://developer.intel.com/quickassist (developers)
- Intel® QuickAssist Technology (QAT) Message of the Week (Full Collateral List)
https://cdrdv2.intel.com/v1/dl/getContent/738128 (requires account)
- Getting Started Guides:
Intel® QuickAssist Technology Software for Linux* - Getting Started Guide - HW version 1.7
https://cdrdv2.intel.com/v1/dl/getContent/710059 (public)
Intel® QuickAssist Technology Software for Linux* - Getting Started Guide - HW version 1.8
https://cdrdv2.intel.com/v1/dl/getContent/729701 (requires account)
Intel® QuickAssist Technology Software for Linux* - Getting Started Guide - HW version 2.0
https://cdrdv2.intel.com/v1/dl/getContent/632506 (requires account)
- Programmer's Guides:
Intel® QuickAssist Technology Software for Linux* - Programmer's Guide - HW version 1.7
https://cdrdv2.intel.com/v1/dl/getContent/710060 (public)
Intel® QuickAssist Technology Software for Linux* - Programmer's Guide - HW version 1.8
https://cdrdv2.intel.com/v1/dl/getContent/613626 (requires account)
Intel® QuickAssist Technology Software for Linux* - Programmer's Guide - HW version 2.0
(TBD)
===============================================================================
Overview
=========
Intel® QuickAssist Technology provides cryptographic and compression acceleration
capabilities used to improve performance and efficiency across the data center.
Note: This README applies to QAT1.7, QAT1.8, and QAT2.0 devices. Not all
sections may apply to your device generation.
===============================================================================
Licensing
==========
The Licensing of the files within this project splits under various set of
licenses which are listed below.
- Dual BSD/GPLv2 License
- BSD License
- GPLv2 License
- OpenSSL License
- ZLIB License
For full GPLv2 License, please see the file LICENSE.GPL contained in the top
level folder.
For full OpenSSL License, please see the file LICENSE.OPENSSL contained in
the top level folder.
For Dual BSD/GPLv2, BSD and ZLIB Licenses, please see the file headers of
the relevant files.
Binary files are provided under the binary redistribution license. Please see
the file binary-redistribution-license.txt contained in the top level folder.
===============================================================================
Compilation and installation
===============================================================================
1) QAT compilation requires the following dependencies:
Libraries:
- libboost >= 1.14
- libudev >= 1.47
- zlib >= 1.2.3
Tools:
- gcc >= 4.4.7
- g++ >= 4.4.7
- pkg-config >= 0.9.0
Development packages:
- yasm
1.1)
- libnl (QAT 1.8 HW)
- isa-l = 2.26 (QAT 2.0 HW)
===============================================================================
2) Configuration process:
To setup the default build configuration - call:
# ./configure
Any other configuration setup can be found under
"configuration options section" or by calling:
# ./configure --help
===============================================================================
3) Compilation process:
After the package has been configured properly,
follow the steps below:
Build and install:
# make
# make install
Build and install sample code:
# make samples
# make samples-install
Uninstall and clean up:
# make uninstall
# make clean
# make distclean
NOTE:
If kernel module signing is required for the QAT out-of-tree drivers,
(e.g. if UEFI Secure Boot is enabled), one can incorporate this step by
setting the 'mod_sign_cmd' environment variable prior to invoking the
'make install' command.
The mod_sign_cmd environment variable should be set to the path of the
required kernel headers sign-file program and include its first three command
line arguments.
e.g.:
# export mod_sign_cmd='<kernel header path>/scripts/sign-file sha256 /root/priv.key /root/pub.der'
===============================================================================
4) Configuration options
These are common examples of the configure command:
# ./configure ICP_ANY_FLAG=path
# ./configure --enable-something
# ./configure ICP_ANY_PATH=path ICP_ANY_NAME=name --enable-something
Options are split into the two following sections: Feature flags and Build flags.
Feature flags:
Enables or disables the additional features supported by QAT package.
--disable-option-checking (QAT 1.7/1.8 HW)
Ignores unrecognized configure options.
--enable-icp-debug
Enables debugging.
Warning: This option should never be enabled in a production environment
as it may output private key information to the console/logs and may also
introduce side channel timing attack vulnerabilities (disabled by default).
--disable-param-check
Disables parameters checking in the top-level APIs.
(Use for performance optimization)
--disable-dc-dyn (QAT 2.0 HW)
Disables Dynamic compression support.
--disable-stats
Disables statistic collection.
(Use for performance optimization)
--enable-icp-log-syslog
Enables debugging messages to be outputted to the
system log instead of standard output.
--enable-icp-nonblocking-partials-perform
Partial operations results are not being blocked.
--enable-icp-sriov
Enables Single-root I/O Virtualization in the QAT driver.
Available options: host, guest.
Example:
./configure --enable-icp-sriov=host
./configure --enable-icp-sriov=guest
--enable-icp-trace
Enables tracing for the Cryptography API.
--enable-icp-asym-only (QAT 2.0 HW)
Enables asymmetric crypto services only.
--enable-icp-sym-only
Enables symmetric crypto services only.
--enable-icp-dc-only
Enables compression service.
(Can optimize size of build objects)
--enable-icp-dc-sym-only
Enables data compression and symmetric crypto services only.
--enable-icp-dc-return-counters-on-error
Enables updates of consumed/produced results in case of
error during compression or decompression operations.
--enable-icp-hb-fail-sim (QAT 1.7/1.8 HW)
Enable Heartbeat Failure Simulation.
--enable-qat-coexistence
Enables legacy and upstream driver coexistence
--enable-qat-lkcf
Enables QAT registration with Linux Kernel Crypto Framework.
--disable-dc-strict-mode (QAT 1.7/1.8 HW)
Disables strict mode for data compression.
Warning: Compressed data might not be de-compressible.
--enable-dc-error-simulation (QAT 1.7/1.8 HW)
Enables Data Compression Error Simulation.
--enable-icp-without-thread (QAT 1.7/1.8 HW)
Removes mutex and spin locks for single thread applications.
--enable-legacy-algorithms (QAT 1.7 and above)
Enable legacy algorithms.
Build flags:
This set of flags have to be set in order to process a proper QAT build.
(Initial values are assigned by default)
If a different setup if required - flags can be easily modified.
Warning.
Please keep in mind that any incorrect change might cause
compilation problems - in such case default configuration setup is
recommended to fix it.
ICP_ROOT
Path to the directory where acceleration software is extracted.
(QAT package root directory)
Type: path
Default value: `pwd`
ICP_BUILDSYSTEM_PATH
Path to build system folder located under the quickassist directory.
Type: path
Default value: `pwd`/quickassist/build_system
ICP_BUILD_OUTPUT
Path to output directory where executables/libraries will be placed.
Type: path
Default value: `pwd`/build
ICP_ENV_DIR
Path to the directory containing the environment build files.
Type: path
Default value: `pwd`/quickassist/build_system/build_files/env_files
ICP_TOOLS_TARGET
Software platform target (Set ‘accelcomp’ for Intel® Communications
Chipset Series Software platforms).
Type: name
Default value: accelcomp
SYSROOT
(QAT 2.0 HW)
Set the logical root directory for headers and libraries. This variable
is used for cross-compilation and is optional.
Type: path
Default value: none
MAX_MR
Number of Miller Rabin rounds for prime operations. Setting this
to a smaller value reduces the memory usage required by the driver.
Type: value
Default value: 50
KERNEL_SOURCE_ROOT
Path to kernel sources directory.
Type: path
Default value: /lib/modules/`uname –r`/build/
===============================================================================
5) Known Issues
The current build system does not support dependency checking functionality
which cause a call of clean during make.
To avoid building the QAT package twice, first during "make" and second during
"make install", it is recommended to invoke only "make install" omitting "make".
# make install
This same issue applies to "make samples" and "make samples-install".
# make samples-install
Legal/Disclaimers
===================
INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL(R) PRODUCTS.
NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL
PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN INTEL'S
TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY
WHATSOEVER, AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO
SALE AND/OR USE OF INTEL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING
TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY
PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. Intel® products are
not intended for use in medical, life saving, life sustaining, critical control
or safety systems, or in nuclear facility applications.
Intel® may make changes to specifications and product descriptions at any time,
without notice.
(C) Intel® Corporation 2022
* Other names and brands may be claimed as the property of others.
===============================================================================
```