CHANGELOG.md

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[0.23.0] - 2024-05-28

Fixed

• Array alignment problem for stack arrays allocated for kernel arguments. (#1357)
• Issue #892, #906 caused by incorrect code generation for indexing (#1377)
• Generation of KernelHasReturnValueError error inside KernelDispatcher. (#1394)
• Issue #1390: broken support for slicing into dpctl.tensor.usm_ndarray in kernels (#1425)
• Support for Wheels package on Windows (#1430)
• Incorrect mangled name for kernel function arguments (#1443)
• Remove artifacts from conda/wheel packages residing in root level (#1450)
• GDB tests to work properly on Intel Max GPU (#1451)
• Improper wheels installation on unsupported platforms (#1452)
• Ref-counting of Python object temporaries in unboxing code (#1454)
• Segfault caused by using malloc to allocate NRT_MemInfo. Replaced with Numba's NRT alloc (#1458)
• Incorrect package name in README.md (#1463)

Added

• A new overloaded dimensions attribute for all index-space id classes (#1359)
• Support for AtomicRef creation using multi-dimensional arrays (#1367)
• Support for linearized indexing functions inside a JIT compiled kernel (#1368)
• Improved documentation: overview (#1341), kernel programming guide (#1388), API docs (#1414), configs options (#1415), comparison with SYCL API (#1417)
• New PrivateArray class in kernel_api to replace dpex.private.array (#1370, #1377)
• Support for libsycinterface::DPCTLKernelArgType enum for specifying type of kernel args instead of hard coding (#1382)
• New indexing unit tests for kernel_api simulator and JIT compiled modes (#1378)
• New unit tests to verify all kernel_api features usable inside device_func (#1391)
• A sycl::local_accessor-like API (kernel_api.LocalAccessor) for numba-dpex kernel (#1331)
• Specialization support for device_func decorator (#1398)
• Support for all kernel_api functions inside the numba_dpex.kernel decorator. (#1400)
• Support for dpnp 0.15 (#1434, #1464)
• Improvements to pyproject.toml configs to build numba-dpex from source. (#1449)
• Load the SPV_INTEL_variable_length_array SPIR-V extension to supporting arrays in private address-space on Intel Max GPU. (#1451)

Changed

• Default inline threshold value set to 2 from None. (#1385)
• Port parfor kernel templates to kernel_api (#1416), (#1424)
• Use SPIRVKernelDispatcher for parfor kernel dispatch (#1435, #1448)
• All examples use the latest dpctl API (#1431)
• Minimum required dpctl version is now 0.16.1
• Minimum required numba version is now 0.59.0 (#1462)

Removed

• OpenCL-like kernel API functions (#1420)
• func decorator (replaced by device_func) (#1400)
• numba_dpex.experimental.kernel and numba_dpex.experimental.device_func (#1400)

[0.22.0] - 2024-02-19

Fixed

• Bug in boxing a DpnpNdArray from parent (#1155)
• Strided layouts and F-contiguous layouts supported in experimental kernel (#1178)
• Barrier call code-generation on OpenCL CPU devices (#1280, #1310)
• Importing numba-dpex can break numba execution (#1267)
• Overhead on launching numba_dpex.kernel functions (#1236)

Added

• Support for dpctl.SyclEvent data type inside dpjit (#1134)
• Support for kernel_api.Range and kernel_api.NdRange inside dpjit (#1148)
• DPEX_OPT: a numba-dpex-specific optimization level config option (#1158)
• Uploading wheels packages to anaconda (#1160)
• flake8 eradicate linter option (#1177)
• Support dpctl.SyclEvent.wait call inside dpjit (#1179)
• Creation of sycl event and queue inside dpjit (#1193, #1190, #1218)
• Experimental kernel dispatcher for kernel compilation (#1178, #1205)
• Added experimental target context for SPIRV codegen (#1213, #1225)
• GDB test cases in public CI (#1209)
• Async kernel submission option (#1219, #1249)
• A new literal type to store IntEnum as Literal types (#1227)
• SYCL-like memory enum classes to the experimental module (#1239)
• call_kernel function to launch kernels (#1260)
• Experimental overloads for an AtomicRef class and fetch_* methods (#1257, #1261)
• New device-specific USMNdArrayModel for USMNdArray and DpnpNdArray types (#1293)
• Experimental atomic load, store and exchange operations (#1297)
• Kernel_api module to simulate kernel functions in pure Python (#1304, #1326)
• Experimental implementation of group barrier operation (#1280)
• Experimental atomic compare_exchange implementation (#1312)
• Experimental group index class (#1310)
• OpenSSF scorecard (#1320)
• Experimental feature index overload methods (#1323)
• Experimental feature group index overload methods (#1330)
• API Documentation for kernel API (#1332)

Changed

• Switch to dpc++ compiler for building numba-dpex (#1210)
• Versioneer and pytest configs into pyproject.toml (#1212)
• numba-dpex can be imported even if no SYCL device is detected by dpctl (#1272)

Removed

• Kernel launch params as lists/tuple. Only Range/NdRange supported (#1251)
• DEFAULT_LOCAL_SIZE global constant (#1291)
• Functions to invoke spirv-tools utilities from spirv_generator (#1292)
• Incomplete vectorize decorator from numba-dpex (#1298)
• Support for Numba 0.57 (#1307)

Deprecated

• OpenCL-like kernel API functions in numba_dpex.ocldecl module

[0.21.4] - 2023-10-12

Fixed

• Remove dead code to silence Coverity errors. (#1163)

[0.21.3] - 2023-09-28

Fixed

• Pin CI conda channels (#1133)
• Mangled kernel name generation (#1112)

Added

• Support tests on single point precision GPUs (#1143)
• Initial work on Coverity scan CI (#1128)
• Python 3.11 support (#1123)
• Security policy (#1117)
• scikit-build to build native extensions (#1107, #1116, #1127, #1139, #1140)

Changed

• Rename helper function to clearly indicate its usage (#1145)
• The data model used by the DpnpNdArray type for kernel functions(#1118)

Removed

• Support for Python 3.8 (#1113)

[0.21.2] - 2023-08-07

Fixed

• Bugs (#1068, #774) in atomic addition caused due to improper floating point atomic emulation. (#1103)

Changed

• Updated documentation and user guides (#1097, #879)

Removed

• Dependency on spirv-tools (#1103, #1108)
• floating point atomic add emulation using atomic_ops.cl (#1103)
• NUMBA_DPEX_ACTIVATE_ATOMICS_FP_NATIVE configuration option (#1103)

[0.21.1] - 2023-07-17

Changed

• Improved support for queue keyword in dpnp array constructor overloads (#1083)
• Improved reduction kernel example (#1089)

Fixed

• Update Itanium CXX ABI Mangler reference (#1080)
• Update sourceware references in docstrings (#1081)
• Typo in error messages of kernel interface (#1082)

[0.21.0] - 2023-06-17

Added

• Support addition and multiplication-based prange reduction loops (#999)
• Proper boxing, unboxing of dpctl.SyclQueue objects inside dpjit decorated functions (#963, #1064)
• Support for queue keyword arguments inside dpnp array constructors in dpjit (#1032)
• Overloads for dpnp array constructors: dpnp.full (#991), dpnp.full_like (#997)
• Support for complex64 and complex128 types as kernel arguments and in parfors (#1033, #1035)
• New config to run the ConstantSizeStaticLocalMemoryPass optionally (#999)
• Support for Numba 0.57 (#1030, #1003, #1002)
• Support for Python 3.11 (#1054)
• Support for SPIRV 1.4 (#1056, #1060)

Changed

• Parfor lowering happens using the kernel pipeline (#996)
• Minimum required Numba version is now 0.57 (#1030)
• Numba monkey patches are moved to numba_dpex.numba_patches (#1030)
• Redesigned unit test suite (#1018, #1017, #1015, #1036, #1037, #1072)

Fixed

• Fix stride computation when unboxing a dpnp array (#1023)
• Using cached queue instead of creating new one on type inference (#946)
• Fixed bug in reduction mul operation for dpjit (#1048)
• Offload of parfor nodes to OpenCL UHD GPU devices (#1074)

Removed

• Support for offloading NumPy-based parfor nodes to SYCL devices (#1041)
• Removed rename_numpy_functions_pass (#1041)
• Dpnp overloads using stubs (#1041, #1025)
• Support for like keyword argument in dpnp array constructor overloads (#1043)
• Support for NumPy arrays as kernel arguments (#1049)
• Kernel argument access specifiers (#1049)
• Support for dpctl.device_context to launch kernels and njit offloading (#1041)

[0.20.1] - 2023-04-07

Added

• Replaced llvm_spirv from oneAPI path by dpcpp-llvm-spirv package.(#979)
• Added Dockerfile and a manual workflow to publish pre-built packages to the repo.(#973)

Fixed

• Fixed default dtype derivation when creating a dpnp.ndarray. (#993)
• Adjusted test_windows step to work with intel-opencl-rt=2023.1.0. (#990)
• Fixed layout in dpnp overload.(#987)
• Handled the case when arraystruct->meminfo is null to close gh-965. (#972)

[0.20.0] - 2023-03-06

Added

• New dpjit decorator supporting dpnp compilation (#887)
• Boxing and unboxing functionality for dpnp.ndarray to numba_dpex (#902)
• New DpexTarget and dispatcher for compiling dpnp using numba-dpex (#887)
• Overload implementation for dpnp.empty (#902)
• Overload implementation for dpnp.empty_like, dpnp.zeros_like and dpnp.ones_like inside dpjit (#928)
• Overload implementation for dpnp.zeros and dpnp.ones inside dpjit (#923)
• Compilation and offload support for dpnp vector style expressions using Numba parfors (#957)
• Compilation of over 70 ufuncs for dpnp inside dpjit (#957)
• Backported the split parfor pass from upstream Numba. (#949)
• Numba type aliases to numba_dpex. (#851)
• Numba prange alias inside numba_dpex. (#957)
• New LRU cache for kernels (#804) and funcs (#877)
• New Range and NdRange classes for kernel submission that follow sycl's range and ndrange syntax. (#888)
• Monkey pacthes to Numba 0.56.4 to support dpnp ufuncs, allocating dpnp arrays (#954)
• New config flag (NUMBA_DPEX_DUMP_KERNEL_LLVM) to dump a kernel's LLVM IR (#924)
• A badge to our gitter chatroom (#919)
• A small script to update copyright headers (#917)
• A new dpexrt_python extension to support USM allocators for Numba NRT_MemInfo (#902)
• Updated examples for kernel API demonstrating compute-follows-data programming model. (#826)

Changed

• CLK_GLOBAL_MEM_FENCE and CLK_LOCAL_MEM_FENCE flags renamed to GLOBAL_MEM_FENCE and LOCAL_MEM_FENCE. (#844)
• Switched from Ubuntu-latest to Ubuntu-20.04 for conda package build (#836)
• Rename USMNdArrayType to USMNdArray (#851)
• Changes to the Numba type to represent dpnp ndarray typess now renamed to DpnpNdarray (#880)
• Improved exceptions and user errors (#804)
• Updated internal API for kernel interface with improved support for __sycl_usm_array_interface__ protocol (#804)
• Pin generated spirv version for kernels to 1.1 (#885)
• Rename DpexContext and DpexTypingContext to DpexKernelTarget and DpexKernelTypingContext (#887)
• Renamed existing dpnp overloads that used stubs to dpnp_stubs_impl.py (#953)
• Dpctl version requirement mismatch is now a warning and not an ImportError (#925)
• Update to versioneer 0.28 (#827)
• Update to dpctl 0.14 (#858)
• Update linters: black to 23.1.0, isort to 5.12.0 (#900)
• License in setup.py to match actual project licensing (#904)

Fixed

• Kernel specialization, compute follows data programming model for kernels (#804)
• Dispatcher/caching rewrite to address performance regression (#912, #896)
• func decorator qualname ambiguation fix (#905)

Removed

• Removes the numpy_usm_shared module from numba_dpex. (#841)
• Removes the usage of llvmlite.llvmpy (#932)

Deprecated

• Support for NumPy arrays as kernel arguments (#804)
• Kernel argument access specifiers (#804)
• Support for dpctl.device_context to launch kernels and njit offloading (#804)
• Dpnp overloads using stubs. (#953)

[0.19.0] - 2022-11-21

Added

• Supported numba0.56. (#818)
• Supported dpnp0.11 and dpctl0.14.
• Added customized exception classes. (#798)

Fixed

• Fixed a crash when calling take() for input array with non-integer values. (#771)
• Fixed pairwise_distance.py to run on machine with no FP64 support in HW. (#806)

[0.18.1] - 2022-08-06

Added

• Implemented support for dpnp.empty() (#728)

Changed

• numba-dppy package is now renamed to numba-dpex.

[0.18.0] - 2022-02-22

Added

• Run coverage in GitHub Actions and upload results to coveralls.io (#621)
• Change black to only allow 80 char lines. Reformat sources. (#631)
• Ignore formatting changes from git-blame. (#632)
• Add numba_support.py with numba_version (#656)
• Add skip_no_numba055 decorator (#662)
• Parameterize test for atomics (#661)
• Reuse decorator skip_no_opencl_Xpu to skip tests (#663)
• Add decorator to skip unsupported atomics (#664)
• Support arrays with __sycl_usm_array_interface__ attribute (#629)
• Support memory allocation in private address space (#640)
• Move skips for opencl to helper (#665)
• Support dpctl 0.12 (#669)
• Implement compute-follows-data programming model [kernel API] (#598)
• Use filter_str to skip tests on missing devices (#672)
• Add check for DPNP and pin MKL version in workflow and dev environment (#648)
• Add CODEOWNERS for distributing review process (#670)
• Add skip_no_dpnp and apply it to all tests (#668)
• Test skipping improvements (#675)
• Use Python 3.9 in dev environment and pin DPNP (#644)
• Add examples into package (#680)
• Make possible to force debugging tests (#681)
• Refactoring for debugging tests (#682)
• Adopt Numba 0.55 debugging features (#654)
• Run public CI on pull request (#695)
• Support for coverage in internal CI (#708)
• Update to dpnp 0.10 (#709)
• Update recipe with dpctl and dpnp version for build (#710)

Changed

• Move dpcpp/llvm-spirv from runtime to testing dependency (#659)

Fixed

• Fix building with DPNP (#674)
• Fix public CI: opencl driver, channel priority, dpctl version (#691)
• Fix codestyle black (#696)
• Fix documentation generation (#697)
• Load dpctl lib on Linux using libDPCTLSyclInterface.so.0 (#707)
• Fix search llvm-spirv if dpcpp compiler package is not installed (#703)
• Pin dpnp version in runtime to allow dev versions of dpnp (#712)

[0.17.4] - 2021-12-02

Changed

• Move dpcpp/llvm-spirv from runtime to testing dependency by @PokhodenkoSA in #659

[0.17.3] - 2021-11-30

Changed

• Use llvm-spirv from dpcpp compiler package by default [cherry picked from #649] (#651)

Fixed

• Enable offloading for numba.njit in dpctl.deveice_context (#630)
• Fix upload conditions for main and release branches (#610)
• Fix DeprecationWarning when use version.parse() [cherry picked from #639] (#642)

[0.17.2] - 2021-11-15

Changed

• Use llvm-spirv from bin-llvm during build for Linux and Windows (#626, #627)

[0.17.1] - 2021-11-10

Changed

• Update clang to icx (#622)

[0.17.0] - 2021-11-03

Added

• Use Python 3.9 [public CI] by @PokhodenkoSA in #574
• Use NUMBA_DPPY_DEBUG for debugging GDB tests by @PokhodenkoSA in #578
• Preliminary support Numba 0.55 (master branch) by @PokhodenkoSA in #583
• Workflow for manually running tests using Numba PRs by @PokhodenkoSA in #586
• Add public CI trigger on tags by @1e-to in #589
• Upload packages for release* branches by @1e-to in #593
• Update to dpctl 0.11 by @PokhodenkoSA in #595
• Update to dpnp 0.9 by @PokhodenkoSA in #599
• Improve the documenatation landing page by @diptorupd in #601
• Clean up README by @diptorupd in #604

Fixed

• Restrict dpctl to 0.10.* for release 0.16 by @1e-to in #590
• Fix upload from release branch by @1e-to in #596
• Unskip tests passing with dpnp 0.9.0rc1 by @PokhodenkoSA in #606

[0.16.1] - 2021-10-20

Changed

• Fix dpctl to 0.10 (#590)
• Add Public CI trigger for tags (#589)

[0.16.0] - 2021-09-28

Added

• Improve build and infra scripts (#544)
• Add docs about local variables lifetime (#534)
• Public CI for Windows (#536, #558)
• Add info about tags in documentation (#543)
• Add code coverage configurations (#561)
• Add support pytest-cov and pytest-xdist (#562)
• Add documentation workflow (#547)
• Test numba and numba-dppy API with GDB (#566)
• Transform commands scripts for GDB to tests (#568)

Changed

• Update dpnp 0.8 (#524)
• Fix passing strides array to DPNP dot and matmul (#565)
• Use older compiler for backwards compatibility (#549)
• Update conda recipe dependency for dpnp (#535)
• Update dpctl 0.10 (memcpy async) (#529)
• Change channels priority in public CI (#532)
• Added runtime dependency llvm-spirv 11.* (#523)
• Update test matrix in README (#560)
• Use dpctl 0.10* and dpnp 0.8* in development configuration (environment.yml)

Fixed

• Update test and fix typo for atomics (#550)
• Delete unused file run_test.sh
• Fix Public CI for using development packages (#522)
• Removed redundant import in docs (#521)

[0.15.0] - 2021-08-25

Added

• Introduce array ultilites to check, allocate and copy memory using SYCL USM (#489)
• Add packaging in run dependencies (#505)
• Add skipping tests for run without GPU (#508)
• Add CI pipeline on GitHub Actions for Linux (#507)
• Enable dpctl.tensor.usm_ndarray for @dppy.kernel (#509)
• Enable @vectorize for target dppy (#497)
• Add integration channels to GitHub Actions and make workflow consistent with dpctl (#510)

Changed

• Update to Numba 0.54.0 (#493)
• Update to dpctl 0.9 (#514)
• Update to dpnp 0.7 (#513)
• Use dpcpp compiler package for Linux (#502)
• Update go version (#515)

Removed

• Remove llvmdev from runtime dependecies (#498)

Fixed

• Fix required compiler flags for processing genreated DWARF symbols (#500)

[0.14.4] - 2021-07-09

Fixed

• Fix emission of debug information (#424)
• Fix examples demonstrating GDB usage on a numba_dppy.kernel function. (#455)
• Remove address space cast from global to generic inside numba_dppy.kernel (#432)
• Fix debugging of local variables (#432)
• Assert offload to SYCL devices in tests (#466)
• Removed calling opt to convert LLVM IR to LLVM Bitcode (#481)

Added

• Add examples for debugging (#426)
• Added a new NUMBA_DPPY_DEBUGINFO environment variable to control GDB usage (#460)
• Add debug option to dppy.kernel decorator (#424)
• Enable debugging of nested GPU functions (#424)
• Enable setting breakpoints by function names while Debugging (#434)
• Various fixes and improvements to documentation about debugging (#479, #474, #475, #480, #475, #477, #468,#450)
• Add automatic generation of commands for debugging (#463)
• Add tests on debugging local variables (#421)
• Enable eager compilation of numba_dppy.kernel (#291)
• Increase test coverage for native atomic ops (#435)
• Check and deter users from returning values from numba_dppy.kernel (#476)

[0.14.3] - 2021-05-27

Fixed

• Add check for ONEAPI_ROOT dir (#411)
• Fix using unquoted environment variable for clang path (#386)
• Fix kernel caching (#408)

[0.14.2] - 2021-05-26

Added

• Update documentation: version 0.14 (#378), API docs (#388), note about Intel Python Numba (#389),
• Update User Guides about Debugging (#380), recommendations (#323), locals (#394), stepping (#400), configure environment (#402), set up machine (#396), info functions (#405)
• Update Developer Guides about DPNP integration (#362)
• Update README: add link to docs (#379), add Cython and pytest in dependencies, add test matrix (#305)
• Add initial integration testing with dpnp and usm_ndarray (#403)
• Introduce type in Numba-dppy to represent dpctl.tensor.usm_ndarray (#391)
• Improve error reporting when searching for dpctl. (#368)
• Enable Python 3.8 in CI (#359)
• Adds a new utils submodule to provide LLVM IR builder helpers. (#355)
• Improve warning and error messages when parfor offload fails. (#353)
• Extend itanium mangler to support numba.types.CPointer and add test (#339)
• Enable optimization level setting (#62)
• Improve message printed during parfor lowering. (#337)
• Initial tests for debug info (#297)
• Add Bandit checks (#264)

Changed

• Update to dpctl 0.8 (#375)
• Update to Numba 0.53 (#279), inluding override get_ufunc_info in DPPYTargetContext (#367)
• Update to DPNP 0.6 (#359)
• Refactor test function generation (#374)
• Ignore the cython generated cpp files inside dpnp_glue. (#351)
• Add automerge main to gold/2021 (#349)
• Fix dpnp version restriction in conda recipe (#347)
• Change all Numba-dppy examples to work wih dpctl 0.7.0 (#309)
• Restrict dpnp version (#344)
• Feature changes related to dpctl 0.7 (#340)
• Rename dpNP to dpnp (#334)
• Ignore generated spir files (#333)
• Use correct names for other products dpctl, Numba, dpnp (#310)
• Remove dead code only if function name is replaced (#303)
• Update license in conda recipe (#350)
• Update blackscholes examples (#377)

Fixed

• Fix dppy_rt extension (#393)
• Update SYCL Filter String (#390)
• Fix atomics (#346)
• Fixes memory leaks in the usage of dpctl C API functions. (#369)
• Fix SPIR-V validation (#354)
• Fix run pre-commit check on main branch
• Fix tests to skip if device is not available (#345)
• Make Test Matrix table smaller in README (#308)
• Fix black action. (#306)
• Fix "subprocess.check_call" for Windows (#269)

[0.13.1] - 2021-03-11

Fixed

• Add spir file to package.
• Do not modify CC env variable during build.
• Use correct version in documentation.

[0.13.0] - 2021-03-02

Added

• Documentation.
• Add support for dpctl.dparray.
• Support NumPy functions via DPNP: random, linalg, transcendental, array ops, array creation.
• Wheels building.
• Using Bandit for finding common security issues in Python code.

Changed

• Start using black code style formatter.
• Build SPIRV code in setup.py.
• Start using pytest for running tests.
• Start using Apache 2.0 license.
• Consistency of file headers.
• Updated to Numba 0.52, dpCtl 0.6 and dpNP 0.5.1.
• Don't create a new copy of a usm shared array data pointers for kernel call.
• Modify test cases and examples to use Level Zero queue.

Fixed

• Fix incorrect import in examples.

[0.12.0] - 2020-12-17

Added

• numba-dppy is a standalone package now. Added setup.py and conda recipe.
• Offload diagnostics.
• Controllable fallback.
• Add flags to generate debug symbols.
• Implementation of np.linalg.eig, np.ndarray.sum, np.ndarray.max, np.ndarray.min, np.ndarray.mean.
• Two new re-write passes to convert NumPy calls into a pseudo numba_dppy call site to allow target-specific overload of NumPy functions. The rewrite passes is a temporary fix till Numba gains support for target-specific overlaods.
• Updated to dpCtl 0.5.* and dpNP 0.4.*

Changed

• The dpnp interface now uses Numba's @overload functionality as opposed to the previous @lower_builtin method.
• Rename DPPL to DPPY.
• Cleaned test code.
• DPPLTestCase replaced with unittest.TestCase.
• All tests and examples use with device_context.
• Config environment variables starts with NUMBA_DPPY_ (i.e. NUMBA_DPPY_SAVE_IR_FILES and NUMBA_DPPY_SPIRV_VAL)
• Remove nested folder dppl in tests.
• No dependency on cffi.

Removed

• The old backup file.

NUMBA Version 0.48.0 + DPPY Version 0.3.0 (June 29, 2020)

This release includes:

• Caching of dppy.kernels which will improve performance.
• Addition of support for Intel Advisor which will help in profiling applications.