Release Intel QPL v1.8.0 · intel/qpl

Added timeout for synchronous wait in case of hardware disconnect. QPL build will also try to autodetect and use target wait-package if available and auto dispatch its use.
Added the --sync_api option to the Benchmarks Framework, allowing the use of the qpl_execute_job API for performance measurements. This is helpful for measuring latency. Without this option, the framework uses a combination of qpl_submit_job and qpl_wait_job APIs.
Experimental logging of IAA execution time is now available in the Benchmarks Framework. To use this feature, build QPL with -DQPL_EXPERIMENTAL_LOG_IAA=ON. This registers timestamps before accelerator submission and after IAA execution, excluding pre- and post-processing overhead. Note: This feature works only in synchronous mode for operations with a single descriptor submission (e.g., Dynamic compression is not supported). To view the percentage of time spent on the accelerator, run the Benchmarks Framework with the --sync_api and --accel_time options.
Added get_max_transfer_size function to hw_dispatcher to handle 2MB limitation on idxd driver size.

Updated requirements.txt to the latest compatible tools required for building Intel(R) QPL documentation locally.
Updated the GoogleTest* submodule to v1.17.0 release. The current QPL test framework is not compatible with previous GoogleTest* versions.
Removed Universally Unique ID library 'uuid' from QPL dependency list.
Published the Code Of Conduct.
Extended clang-tidy configuration file by QPL headers and refactored it correspondently.
Extended fuzzy testing by acceleration path functionality.
Expanded bad argument and unit testing for the qpl_get_safe_deflate_compression_buffer_size and for the qpl_{deflate, huffman_only}_table_create functions.

Deprecated the sanitizer build options -DSANITIZE_MEMORY and -DSANITTIZE_THREAD. Instead, use -DQPL_USE_CLANG_SANITIZER=[ADDRESS|LEAK|UNDEFINED|THREAD|ALL_COMPATIBLE] to enable sanitizers. This change aligns with Clang sanitizer build options. See the Build Options for more details.

Fixed an integer overflow issue in the qpl_get_safe_deflate_compression_buffer_size function. Previously, for sizes close to UINT32_MAX, users could receive a negative value.
The behavior of QPL dispatcher was changed under Windows - instead of automatic changing qpl_path_hardware to qpl_path_software in qpl_init_job function(accelerator offload is not supported under Windows), accelerator support error is now returning.

Intel(R) QPL could be built from directly downloadable files (.tar, .tgz) without tests and benchmark frameworks, using the -DQPL_BUILD_TESTS=OFF build option. This is because it requires submodules that are not included in the archives by GitHub* during release creation.
Known test failures are listed below. Some tests fail only under certain conditions, which are noted in parentheses.
- Functional tests:
  - (software_path, auto_path only on platforms without IAA) ta_c_api_deflate_stateful.{dynamic/fixed/static}_default_verify
  - (software_path, auto_path) ta_c_api_deflate_stateful.{dynamic/fixed/static}_high_verify
Compression verification on the qpl_path_software works only with indexing mode and data of size smaller than 32KB in other modes.
Inflate does not report the error code QPL_STS_BIG_HEADER_ERR when a header is too big to fit in the input buffer.
The implementation of QPL_FLAG_CRC32C is in progress.
When using qpl_path_hardware, the compression and decompression with indexing mode on IAA 2.0 are limited to data sizes smaller than 4KB.
The Force Array Output Modification feature is enabled only for qpl_path_hardware and Intel IAA 2.0 (and later). In the case of qpl_path_auto, an error code QPL_STS_NOT_SUPPORTED is returned as no fallback is available currently.

The release includes contributions from the project team.

Provide feedback