Version 0.59.0 (31 January 2024) -------------------------------- .. contents:: Table of Contents :depth: 2 This is a major Numba release. Numba now supports Python 3.12, please find a summary of all noteworthy items below. Highlights ~~~~~~~~~~ Python 3.12 Support =================== The standout feature of this release is the official support for Python 3.12 in Numba. Please note that profiling support is temporarily disabled in this release (for Python 3.12) and several known issues have been identified during development. The Numba team is actively working on resolving them. Please refer to the respective issue pages (`Numba #9289 `_ and `Numba #9291 `_) for a list of ongoing issues and updates on progress. (`PR-#9246 `__) Move minimum supported Python version to 3.9. ============================================= Support for Python 3.8 has been removed, Numba's minimum supported Python version is now Python 3.9. (`PR-#9310 `__) New Features ~~~~~~~~~~~~ Add support for ufunc attributes and reduce =========================================== Support for ``ufunc.reduce`` and most ufunc attributes is added. (`PR-#9123 `__) Add a config variable to enable / disable the llvmlite memory manager ===================================================================== A config variable to force enable or disable the llvmlite memory manager is added. (`PR-#9341 `__) Improvements ~~~~~~~~~~~~ Add ``TargetLibraryInfo`` pass to CPU LLVM pipeline. ==================================================== The ``TargetLibraryInfo`` pass makes sure that the optimisations that take place during call simplification are appropriate for the target, without this the target is assumed to be Linux and code will be optimised to produce e.g. math symbols that do not exit on Windows. Historically this issue has been avoided through the use of Numba internal libraries carrying wrapped symbols, but doing so potentially detriments performance. As a result of this change Numba internal libraries are smaller and there is an increase in optimisation opportunity in code using ``exp2`` and ``log2`` functions. (`PR-#9336 `__) Numba deprecation warning classes are now subclasses of builtin ones ==================================================================== To help users manage and suppress deprecation warnings from Numba, the ``NumbaDeprecationWarning`` and ``NumbaPendingDeprecationWarning`` classes are now subclasses of the builtin ``DeprecationWarning`` and ``PendingDeprecationWarning`` respectively. Therefore, warning filters on ``DeprecationWarning`` and ``PendingDeprecationWarning`` will apply to Numba deprecation warnings. (`PR-#9347 `__) NumPy Support ~~~~~~~~~~~~~ Added support for np.indices() function. ======================================== Support is added for ``numpy.indices()``. (`PR-#9126 `__) Added support for ``np.polynomial.polynomial.Polynomial`` class. ================================================================ Support is added for the `Polynomial` class from the package ``np.polynomial.polynomial``. (`PR-#9140 `__) Added support for functions ``np.polynomial.polyutils.as_series()``, as well as functions ``polydiv()``, ``polyint()``, ``polyval()`` from ``np.polynomial.polynomial``. ======================================================================================================================================================================== Support is added for ``np.polynomial.polyutils.as_series()``, ``np.polynomial.polynomial.polydiv()``, ``np.polynomial.polynomial.polyint()`` (only the first 2 arguments), ``np.polynomial.polynomial.polyval()`` (only the first 2 arguments). (`PR-#9141 `__) Added support for np.unwrap() function. ======================================= Support is added for ``numpy.unwrap()``. The ``axis`` argument is only supported when its value equals -1. (`PR-#9154 `__) Adds support for checking if dtypes are equal. ============================================== Support is added for checking if two dtype objects are equal, for example ``assert X.dtype == np.dtype(np.float64)``. (`PR-#9249 `__) CUDA API Changes ~~~~~~~~~~~~~~~~ Added support for compiling device functions with a C ABI ========================================================= Support for compiling device functions with a C ABI through the :func:`compile_ptx() ` API, for easier interoperability with CUDA C/C++ and other languages. (`PR-#9223 `__) Make grid() and gridsize() use 64-bit integers ============================================== ``cuda.grid()`` and ``cuda.gridsize()`` now use 64-bit integers, so they no longer overflow when the grid contains more than ``2 ** 31`` threads. (`PR-#9235 `__) Prevent kernels being dropped by implementing the used list =========================================================== Kernels are no longer dropped when being compiled and linked using nvJitLink, because they are added to the ``@"llvm.used"`` list. (`PR-#9267 `__) Support for Windows CUDA 12.0 toolkit conda packages ==================================================== The library paths used in CUDA toolkit 12.0 conda packages on Windows are added to the search paths used when detecting CUDA libraries. (`PR-#9279 `__) Performance Improvements and Changes ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Improvement to IR copying speed =============================== Improvements were made to the deepcopying of ``FunctionIR``. In one case, the ``InlineInlineables`` pass is 3x faster. (`PR-#9245 `__) Bug Fixes ~~~~~~~~~ Dynamically Allocate Parfor Schedules ===================================== This PR fixes an issue where a parallel region is executed in a loop many times. The previous code used an alloca to allocate the parfor schedule on the stack but if there are many such parfors in a loop then the stack will overflow. The new code does a pair of allocation/deallocation calls into the Numba parallel runtime before and after the parallel region respectively. At the moment, these calls redirect to malloc/free although other mechanisms such as pooling are possible and may be implemented later. This PR also adds a warning in cases where a prange loop is not converted to a parfor. This can happen if there is exceptional control flow in the loop. These are related in that the original issue had a prange loop that wasn't converted to a parfor and therefore all the parfors inside the body of the prange were running in parallel and adding to the stack each time. (`PR-#9048 `__) Support multiple outputs in a ``@guvectorize`` function ======================================================= This PR fixes `Numba #9058 `_ where it is now possible to call a guvectorize with multiple outputs. (`PR-#9049 `__) Handling of ``None`` args fixed in ``PythonAPI.call``. ====================================================== Fixing segfault when ``args=None`` was passed to ``PythonAPI.call``. (`PR-#9089 `__) Fix propagation of literal values in PHI nodes. =============================================== Fixed a bug in the literal propagation pass where a PHI node could be wrongly replaced by a constant. (`PR-#9144 `__) ``numpy.digitize`` implementation behaviour aligned with numpy ============================================================== The implementation of ``numpy.digitize`` is updated to behave per numpy in a wider set of cases, including where the supplied bins are not in fact monotonic. (`PR-#9169 `__) ``numpy.searchsorted`` and ``numpy.sort`` behaviour updates =========================================================== * ``numpy.searchsorted`` implementation updated to produce identical outputs to numpy for a wider set of use cases, including where the provided array `a` is in fact not properly sorted. * ``numpy.searchsorted`` implementation bugfix for the case where side='right' and the provided array `a` contains NaN(s). * ``numpy.searchsorted`` implementation extended to support complex inputs. * ``numpy.sort`` (and ``array.sort``) implementation extended to support sorting of complex data. (`PR-#9189 `__) Fix SSA to consider variables where use is not dominated by the definition ========================================================================== A SSA problem is fixed such that a conditionally defined variable will receive a phi node showing that there is a path where the variable is undefined. This affects extension code that relies on SSA behavior. (`PR-#9242 `__) Fixed ``RecursionError`` in ``prange`` ====================================== A problem with certain loop patterns using ``prange`` leading to ``RecursionError`` in the compiler is fixed. An example of such loop is shown below. The problem would cause the compiler to fall into an infinite recursive cycle trying to determine the definition of ``var1`` and ``var2``. The pattern involves definitions of variables within an if-else tree and not all branches are defining the variables. .. code-block:: for i in prange(N): for j in inner: if cond1: var1 = ... elif cond2: var1, var2 = ... elif cond3: pass if cond4: use(var1) use(var2) (`PR-#9244 `__) Support negative axis in ufunc.reduce ===================================== Fixed a bug in ufunc.reduce to correctly handle negative axis values. (`PR-#9296 `__) Fix issue with parfor reductions and Python 3.12. ================================================= The parfor reduction code has certain expectations on the order of statements that it discovers, these are based on the code that previous versions of Numba generated. With Python 3.12, one assignment that used to follow the reduction operator statement, such as a binop, is now moved to its own basic block. This change reorders the set of discovered reduction nodes so that this assignment is right after the reduction operator as it was in previous Numba versions. This only affects internal parfor reduction code and doesn't actually change the Numba IR. (`PR-#9334 `__) Changes ~~~~~~~ Make test listing not invoke CPU compilation. ============================================= Numba's test listing command ``python -m numba.runtests -l`` has historically triggered CPU target compilation due to the way in which certain test functions were declared within the test suite. It has now been made such that the CPU target compiler is not invoked on test listing and a test is added to ensure that it remains the case. (`PR-#9309 `__) Semantic differences due to Python 3.12 variable shadowing in comprehensions ============================================================================ Python 3.12 introduced a new bytecode ``LOAD_FAST_AND_CLEAR`` that is only used in comprehensions. It has dynamic semantics that Numba cannot model. For example, .. code-block:: python def foo(): if False: x = 1 [x for x in (1,)] return x # This return uses undefined variable The variable `x` is undefined at the return statement. Instead of raising an ``UnboundLocalError``, Numba will raise a ``TypingError`` at compile time if an undefined variable is used. However, Numba cannot always detect undefined variables. For example, .. code-block:: python def foo(a): [x for x in (0,)] if a: x = 3 + a x += 10 return x Calling ``foo(0)`` returns ``10`` instead of raising ``UnboundLocalError``. This is because Numba does not track variable liveness at runtime. The return value is ``0 + 10`` since Numba zero-initializes undefined variables. (`PR-#9315 `__) Refactor and remove legacy APIs/testing internals. ================================================== A number of internally used functions have been removed to aid with general maintenance by reducing the number of ways in which it is possible to invoke compilation, specifically: * ``numba.core.compiler.compile_isolated`` is removed. * ``numba.tests.support.TestCase::run_nullary_func`` is removed. * ``numba.tests.support.CompilationCache`` is removed. Additionally, the concept of "nested context" is removed from ``numba.core.registry.CPUTarget`` along with the implementation details. Maintainers of target extensions (those using the API in ``numba.core.target_extension`` to extend Numba support to custom/synthetic hardware) should note that the same can be deleted from target extension implementations of ``numba.core.descriptor.TargetDescriptor`` if it is present. i.e. the ``nested_context`` method and associated implementation details can just be removed from the custom target's ``TargetDescriptor``. Further, a bug was discovered, during the refactoring, in the typing of record arrays. It materialised that two record types that only differed in their mutability could alias, this has now been fixed. (`PR-#9330 `__) Deprecations ~~~~~~~~~~~~ Explicitly setting ``NUMBA_CAPTURED_ERRORS=old_style`` will raise deprecation warnings ====================================================================================== As per deprecation schedule of old-style error-capturing, explicitly setting ``NUMBA_CAPTURED_ERRORS=old_style`` will raise deprecation warnings. This release is the last to use "old_style" as the default. Details are documented at https://numba.readthedocs.io/en/0.58.1/reference/deprecation.html#deprecation-of-old-style-numba-captured-errors (`PR-#9346 `__) Expired Deprecations ~~~~~~~~~~~~~~~~~~~~ Object mode *fall-back* support has been removed. ================================================= As per the deprecation schedule for Numba 0.59.0, support for "object mode fall-back" is removed from all Numba ``jit``-family decorators. Further, the default for the ``nopython`` key-word argument has been changed to ``True``, this means that all Numba ``jit``-family decorated functions will now compile in ``nopython`` mode by default. (`PR-#9352 `__) Removal of deprecated API ``@numba.generated_jit``. =================================================== As per the deprecation schedule for 0.59.0, support for ``@numba.generated_jit`` has been removed. Use of ``@numba.extending.overload`` and the high-level extension API is recommended as a replacement. (`PR-#9353 `__) Infrastructure Related Changes ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Add validation capability for user generated towncrier ``.rst`` files. ====================================================================== Added a validation script for user generated towncrier ``.rst`` files. The script will run as a part of towncrier Github workflow automatically on every PR. (`PR-#9335 `__) Pull-Requests: ~~~~~~~~~~~~~~ * PR `#8990 `_: Removed extra block copying in InlineWorker (`kc611 `_) * PR `#9048 `_: Dynamically allocate parfor schedule. (`DrTodd13 `_) * PR `#9058 `_: Fix gufunc with multiple outputs (`guilhermeleobas `_) * PR `#9089 `_: Fix segfault on passing `None` for args in PythonAPI.call (`hellozee `_) * PR `#9101 `_: Add misc script to find missing towncrier news files (`sklam `_) * PR `#9123 `_: Implement most ufunc attributes and ufunc.reduce (`guilhermeleobas `_) * PR `#9126 `_: Add support for np.indices() (`KrisMinchev `_) * PR `#9140 `_: Add support for Polynomial class (`KrisMinchev `_) * PR `#9141 `_: Add support for `as_series()` from `np.polynomial.polyutils` and `polydiv()`, `polyint()`, `polyval()` from `np.polynomial.polynomial` (`KrisMinchev `_) * PR `#9142 `_: Removed out of date comment handled by PR#8338 (`njriasan `_) * PR `#9144 `_: Fix error when literal is wrongly propagated in a PHI node (`guilhermeleobas `_) * PR `#9148 `_: bump llvmdev dependency to 0.42.0dev for next development cycle (`esc `_) * PR `#9149 `_: update release checklist post 0.58.0rc1 (`esc `_) * PR `#9154 `_: Add support for np.unwrap() (`KrisMinchev `_) * PR `#9155 `_: Remove unused test.cmd (`sklam `_) * PR `#9168 `_: fix the `get_template_info` method in `overload_method` template (`dlee992 `_ `sklam `_) * PR `#9169 `_: Update `np.digitize` handling of np.nan bin edge(s) (`rjenc29 `_) * PR `#9170 `_: Fix an inappropriate test expression to remove a logical short circuit (`munahaf `_) * PR `#9171 `_: Fix the implementation of a special method (`munahaf `_) * PR `#9189 `_: Align `searchsorted` behaviour with numpy (`rjenc29 `_) * PR `#9191 `_: Add a Numba power-on-self-test script and use in CI. (`stuartarchibald `_) * PR `#9205 `_: release notes and version support updates from release0.58 branch (`esc `_) * PR `#9223 `_: CUDA: Add support for compiling device functions with C ABI (`gmarkall `_) * PR `#9235 `_: CUDA: Make `grid()` and `gridsize()` use 64-bit integers (`gmarkall `_) * PR `#9236 `_: Fixes numba/numba#9234 (`SridharCR `_) * PR `#9244 `_: Fix Recursion error in parfor lookup (`sklam `_) * PR `#9245 `_: Fix slow InlineInlineable (`sklam `_) * PR `#9246 `_: Support for Python 3.12 (`stuartarchibald `_ `kc611 `_ `esc `_) * PR `#9249 `_: add support for checking dtypes equal (`saulshanabrook `_) * PR `#9255 `_: Fix SSA to consider variables whose use is not dominated by the definition (`sklam `_) * PR `#9258 `_: [docs] Typo in overloading-guide.rst (`kinow `_) * PR `#9267 `_: CUDA: Fix dropping of kernels by nvjitlink, by implementing the used list (`gmarkall `_) * PR `#9279 `_: CUDA: Add support for CUDA 12.0 Windows conda packages (`gmarkall `_) * PR `#9292 `_: CUDA: Switch cooperative groups to use overloads (`gmarkall `_) * PR `#9296 `_: Fix bug when axis is negative and check when axis is invalid (`guilhermeleobas `_) * PR `#9301 `_: Release Notes 0.58.1 for main (`esc `_) * PR `#9302 `_: add missing backtick to example git tag command (`esc `_) * PR `#9303 `_: Add category to warning (`kkokkoros `_) * PR `#9307 `_: Upgrade to cloudpickle 3.0.0 (`sklam `_) * PR `#9308 `_: Fix typo in azure ci script (`sklam `_) * PR `#9309 `_: Continue #9044, prevent compilation on the CPU target when listing tests. (`stuartarchibald `_ `apmasell `_) * PR `#9310 `_: Remove Python 3.8 support. (`stuartarchibald `_) * PR `#9315 `_: Fix support for LOAD_FAST_AND_CLEAR (`sklam `_) * PR `#9318 `_: GPU CI: Test with Python 3.9-3.12 (`gmarkall `_) * PR `#9325 `_: Fix GPUCI (`gmarkall `_) * PR `#9326 `_: Add docs for LOAD_FAST_AND_CLEAR changes (`sklam `_) * PR `#9330 `_: Refactor and remove legacy APIs/testing internals. (`stuartarchibald `_) * PR `#9331 `_: Fix Syntax and Deprecation Warnings from 3.12. (`stuartarchibald `_) * PR `#9334 `_: Fix parfor reduction issue with Python 3.12. (`DrTodd13 `_) * PR `#9335 `_: Add validation capability for user generated towncrier .rst files. (`kc611 `_) * PR `#9336 `_: Add TargetLibraryInfo pass to CPU LLVM pipeline. (`stuartarchibald `_) * PR `#9337 `_: Revert #8583 which skip tests due to M1 RuntimeDyLd Assertion error (`sklam `_) * PR `#9341 `_: Add configuration variable to force llvmlite memory manager on / off (`gmarkall `_) * PR `#9342 `_: Fix flake8 checks for v6.1.0 (`gmarkall `_) * PR `#9346 `_: Setting ``NUMBA_CAPTURED_ERRORS=old_style`` will now raise warnings. (`sklam `_) * PR `#9347 `_: Make Numba's deprecation warnings subclasses of the builtin ones. (`sklam `_) * PR `#9351 `_: Made Python 3.12 support rst note more verbose (`kc611 `_) * PR `#9352 `_: Removing object mode fallback from `@jit`. (`stuartarchibald `_) * PR `#9353 `_: Remove `numba.generated_jit` (`stuartarchibald `_) * PR `#9356 `_: Refactor print tests to avoid NRT leak issue. (`stuartarchibald `_) * PR `#9357 `_: Fix a typo in `_set_init_process_lock` warning. (`stuartarchibald `_) * PR `#9358 `_: Remove note about OpenMP restriction in wheels. (`stuartarchibald `_) * PR `#9359 `_: Fix test_jit_module test against objmode fallback. (`stuartarchibald `_) * PR `#9360 `_: AzureCI changes. RVSDG test config should still test its assigned test slice (`sklam `_) * PR `#9362 `_: Fix ``np.MachAr`` warning matching in test. (`sklam `_) * PR `#9402 `_: Doc updates for 0.59 final (`sklam `_ `stuartarchibald `_) * PR `#9403 `_: Fix test isolation for stateful configurations in the testsuite (`sklam `_ `stuartarchibald `_) * PR `#9404 `_: Fix skipped test stderr change for Python 3.12.1. (`stuartarchibald `_) * PR `#9407 `_: Fix incorrect cycle detection (`sklam `_) Authors: ~~~~~~~~ * `apmasell `_ * `dlee992 `_ * `DrTodd13 `_ * `esc `_ * `gmarkall `_ * `guilhermeleobas `_ * `hellozee `_ * `kc611 `_ * `kinow `_ * `kkokkoros `_ * `KrisMinchev `_ * `munahaf `_ * `njriasan `_ * `rjenc29 `_ * `saulshanabrook `_ * `sklam `_ * `SridharCR `_ * `stuartarchibald `_