Version 0.60.0 (13 June 2024)
=============================

This is a major Numba release. Numba now has binary support for
NumPy 2.0. Users should note that this does NOT yet include NEP 50
related type-level changes which are still in progress. This
release does not guarantee execution level compatibility with
NumPy 2.0 and hence users should expect some type and numerical
deviations with respect to normal Python behavior while using
Numba with NumPy 2.0.

Please find a summary of all noteworthy items below.

Highlights
~~~~~~~~~~

NumPy 2.0 Binary Support
------------------------

Added Binary Support for NumPy 2.0. However, this does not yet include
NEP 50 related type-level changes which are still in progress.

Following is a summary of the user facing changes:

* The ``ptp()`` method previously available for arrays has been deprecated.
  Instead, it is recommended to use the ``np.ptp(arr)`` function.
* The data type ``np.bool8`` has been deprecated and replaced with ``np.bool``.
* The ``np.product`` function is deprecated; users are advised to use
  ``np.prod`` instead.
* Starting from NumPy version 2.0, the ``itemset()`` method has been removed
  from the ``ndarray`` class. To achieve the same functionality, utilize
  the assignment operation ``arr[index] = value``.
* Deprecated constants ``np.PINF`` and ``np.NINF`` should be replaced with
  ``np.inf`` for positive infinity and ``-np.inf`` for negative infinity,
  respectively.

(`PR-#9466 <https://github.com/numba/numba/pull/9466>`__)


New Features
~~~~~~~~~~~~

Enhance guvectorize support in JIT code
---------------------------------------

Generalized universal function support is extended, it is now possible to call
a ``@guvectorize`` decorated function from within a JIT-compiled function.
However, please note that broadcasting is not supported yet. Calling a
guvectorize function in a scenario where broadcast is needed may result in
incorrect behavior.

(`PR-#8984 <https://github.com/numba/numba/pull/8984>`__)

Add experimental support for ufunc.at
-------------------------------------

Experimental support for ``ufunc.at`` is added.

(`PR-#9239 <https://github.com/numba/numba/pull/9239>`__)

Add ``float(<string literal>)`` ctor
------------------------------------

Support for ``float(<string literal>)`` is added.

(`PR-#9378 <https://github.com/numba/numba/pull/9378>`__)

Add support for ``math.log2``.
------------------------------

Support for ``math.log2`` is added.

(`PR-#9416 <https://github.com/numba/numba/pull/9416>`__)

Add math.nextafter support for nopython mode.
---------------------------------------------

Support ``math.nextafter`` in nopython mode.

(`PR-#9438 <https://github.com/numba/numba/pull/9438>`__)

Add support for parfor binop reductions.
----------------------------------------

Previously, only operations with inplace operations like `+=` could be used as reductions
in `prange`s.  Now, with this PR, binop reductions of the form `a = a binop b` can be used.

(`PR-#9521 <https://github.com/numba/numba/pull/9521>`__)


Improvements
~~~~~~~~~~~~

Expand ``isinstance()`` support for NumPy datetime types
--------------------------------------------------------

Adds support of ``numpy.datetime64`` and ``numpy.timedelta64`` types in
``isinstance()``.

(`PR-#9455 <https://github.com/numba/numba/pull/9455>`__)

Python 3.12 ``sys.monitoring`` support is added to Numba's dispatcher.
----------------------------------------------------------------------

Python 3.12 introduced a new module ``sys.monitoring`` that makes available an
event driven monitoring API for use in tools that need to monitor execution e.g.
debuggers or profilers. Numba's dispatcher class (the code that handles transfer
of control between the Python interpreter and compiled code) has been updated to
emit ``sys.monitoring.events.PY_START`` and ``sys.monitoring.events.PY_RETURN``
as appropriate. This allows tools that are watching for these events to identify
when control has entered and returned from compiled code. As a result of this
change, Numba compiled code is now identified by ``cProfile`` in the same way
that it has been historically i.e. it will be present in performance profiles.

(`PR-#9482 <https://github.com/numba/numba/pull/9482>`__)


NumPy Support
~~~~~~~~~~~~~

Added support for ``np.size()``
-------------------------------

Added ``np.size()`` support for NumPy, which was previously unsupported.

(`PR-#9504 <https://github.com/numba/numba/pull/9504>`__)


CUDA API Changes
~~~~~~~~~~~~~~~~

Support for compilation to LTO-IR
---------------------------------

Support for compiling device functions to LTO-IR in the compilation API is
added.

(`PR-#9274 <https://github.com/numba/numba/pull/9274>`__)

Support math.log, math.log2 and math.log10 in CUDA
--------------------------------------------------

CUDA target now supports ``np.log``, ``np.log2`` and ``np.log10``.

(`PR-#9417 <https://github.com/numba/numba/pull/9417>`__)


Bug Fixes
~~~~~~~~~

Fix parfor variable hoisting analysis.
--------------------------------------

If a variable is used to build a container (e.g., tuple, list, map, set) or is passed as an
argument to a call then conservatively assume it could escape the current iteration of
the parfor and so should not be hoisted.

(`PR-#9532 <https://github.com/numba/numba/pull/9532>`__)


Deprecations
~~~~~~~~~~~~

Deprecate `old_style` error-capturing
-------------------------------------

Per deprecation schedule, `old_style` error-capturing is deprecated and the
`default` is now `new_style`.

(`PR-#9549 <https://github.com/numba/numba/pull/9549>`__)


Expired Deprecations
~~~~~~~~~~~~~~~~~~~~

Removal of ``numba.core.retarget``
----------------------------------

The experimental features implemented in ``numba.core.retarget`` have been
removed. These features were primarily used in numba-dpex, but that project has
replaced its use of ``numba.core.retarget`` with a preference for
*target extension API*.

(`PR-#9539 <https://github.com/numba/numba/pull/9539>`__)


Documentation Changes
~~~~~~~~~~~~~~~~~~~~~

``numba.cuda.gpus.current`` documentation correction
----------------------------------------------------

``numba.cuda.gpus.current`` was erroneously described
as a function, is now described as an attribute.

(`PR-#9394 <https://github.com/numba/numba/pull/9394>`__)

CUDA 12 conda installation documentation
----------------------------------------

Installation instructions have been added for CUDA 12 conda users.

(`PR-#9487 <https://github.com/numba/numba/pull/9487>`__)

Version 0.59.1 (18 March 2024)
------------------------------

This is a bug-fix release to fix regressions in 0.59.0.


CUDA API Changes
~~~~~~~~~~~~~~~~

Fixed caching of kernels that use target-specific overloads
===========================================================

Caching of kernels using target-specific overloads now works. This includes use
of cooperative group sync, which is now implemented with a target-specific overload.

(`PR-#9447 <https://github.com/numba/numba/pull/9447>`__)


Performance Improvements and Changes
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Improvement to ``np.searchsorted``
==================================

Fixed a performance regression introduced in Numba 0.59 which made
``np.searchsorted`` considerably slower.

(`PR-#9448 <https://github.com/numba/numba/pull/9448>`__)


Bug Fixes
~~~~~~~~~

Fix issues with ``np.searchsorted`` not handling ``np.datetime64``
==================================================================

This patch fixes two issues with ``np.searchsorted``. First, a regression is
fixed in the support of ``np.datetime64``. Second, adopt ``NAT``-aware
comparisons to fix mishandling of ``NAT`` value.

(`PR-#9445 <https://github.com/numba/numba/pull/9445>`__)

Allow use of Python 3.12 PEP-695 type parameter syntax
======================================================

A patch is added to properly parse the PEP 695 syntax. While Numba
does not yet take advantage of type parameters, it will no longer erroneously
reject functions defined with the new Python 3.12 syntax.

(`PR-#9459 <https://github.com/numba/numba/pull/9459>`__)


Version 0.59.0 (31 January 2024)
--------------------------------

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 <https://github.com/numba/numba/pull/9289>`_ and
`Numba #9291 <https://github.com/numba/numba/pull/9291>`_)
for a list of ongoing issues and updates on progress.

(`PR-#9246 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/9310>`__)


New Features
~~~~~~~~~~~~

Add support for ufunc attributes and reduce
===========================================

Support for ``ufunc.reduce`` and most ufunc attributes is added.

(`PR-#9123 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/9347>`__)


NumPy Support
~~~~~~~~~~~~~

Added support for np.indices() function.
========================================

Support is added for ``numpy.indices()``.

(`PR-#9126 <https://github.com/numba/numba/pull/9126>`__)

Added support for ``np.polynomial.polynomial.Polynomial`` class.
================================================================

Support is added for the `Polynomial` class from the package ``np.polynomial.polynomial``.

(`PR-#9140 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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() <numba.cuda.compile_ptx>` API, for easier interoperability
with CUDA C/C++ and other languages.

(`PR-#9223 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/9048>`__)

Support multiple outputs in a ``@guvectorize`` function
=======================================================

This PR fixes `Numba #9058 <https://github.com/numba/numba/pull/9058>`_
where it is now possible to call a guvectorize with multiple outputs.

(`PR-#9049 <https://github.com/numba/numba/pull/9049>`__)

Handling of ``None`` args fixed in ``PythonAPI.call``.
======================================================

Fixing segfault when ``args=None`` was passed to ``PythonAPI.call``.

(`PR-#9089 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/9244>`__)

Support negative axis in ufunc.reduce
=====================================

Fixed a bug in ufunc.reduce to correctly handle negative axis values.

(`PR-#9296 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/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 <https://github.com/numba/numba/pull/9335>`__)