From d10576f4cafd51fba3f19d2654ddb0a1c5f2fde0 Mon Sep 17 00:00:00 2001 From: Stuart Archibald Date: Thu, 27 Jul 2023 15:46:45 +0100 Subject: [PATCH 1/2] Add towncrier news snippets for PRs that are missing them. This adds news snippets for PRs that are scheduled for Numba 0.58 but were merged prior to towncrier being enabled and a news snippet being a requirement. Closes #9070 --- docs/upcoming_changes/7262.bug_fix.rst | 5 +++++ docs/upcoming_changes/7543.np_support.rst | 5 +++++ docs/upcoming_changes/8462.improvement.rst | 6 ++++++ docs/upcoming_changes/8736.np_support.rst | 5 +++++ docs/upcoming_changes/8885.bug_fix.rst | 5 +++++ docs/upcoming_changes/8892.new_feature.rst | 8 ++++++++ docs/upcoming_changes/8906.new_feature.rst | 17 +++++++++++++++++ docs/upcoming_changes/8911.improvement.rst | 5 +++++ docs/upcoming_changes/8916.change.rst | 4 ++++ docs/upcoming_changes/8944.bug_fix.rst | 4 ++++ docs/upcoming_changes/8974.cuda.rst | 13 +++++++++++++ docs/upcoming_changes/8988.cuda.rst | 4 ++++ docs/upcoming_changes/8995.np_support.rst | 7 +++++++ docs/upcoming_changes/9035.cuda.rst | 6 ++++++ 14 files changed, 94 insertions(+) create mode 100644 docs/upcoming_changes/7262.bug_fix.rst create mode 100644 docs/upcoming_changes/7543.np_support.rst create mode 100644 docs/upcoming_changes/8462.improvement.rst create mode 100644 docs/upcoming_changes/8736.np_support.rst create mode 100644 docs/upcoming_changes/8885.bug_fix.rst create mode 100644 docs/upcoming_changes/8892.new_feature.rst create mode 100644 docs/upcoming_changes/8906.new_feature.rst create mode 100644 docs/upcoming_changes/8911.improvement.rst create mode 100644 docs/upcoming_changes/8916.change.rst create mode 100644 docs/upcoming_changes/8944.bug_fix.rst create mode 100644 docs/upcoming_changes/8974.cuda.rst create mode 100644 docs/upcoming_changes/8988.cuda.rst create mode 100644 docs/upcoming_changes/8995.np_support.rst create mode 100644 docs/upcoming_changes/9035.cuda.rst diff --git a/docs/upcoming_changes/7262.bug_fix.rst b/docs/upcoming_changes/7262.bug_fix.rst new file mode 100644 index 00000000000..1b4b53f00ab --- /dev/null +++ b/docs/upcoming_changes/7262.bug_fix.rst @@ -0,0 +1,5 @@ +Handling of different sized unsigned integer indexes are fixed in ``numba.typed.List``. +======================================================================================= + +An issue with the order of truncation/extension and casting of unsigned integer +indexes in ``numba.typed.List`` has been fixed. diff --git a/docs/upcoming_changes/7543.np_support.rst b/docs/upcoming_changes/7543.np_support.rst new file mode 100644 index 00000000000..578bfebd7fb --- /dev/null +++ b/docs/upcoming_changes/7543.np_support.rst @@ -0,0 +1,5 @@ +All modes are supported in ``numpy.correlate`` and ``numpy.convolve``. +====================================================================== + +All values for the ``mode`` argument to ``numpy.correlate`` and +``numpy.convolve`` are now supported. diff --git a/docs/upcoming_changes/8462.improvement.rst b/docs/upcoming_changes/8462.improvement.rst new file mode 100644 index 00000000000..4f03ca06818 --- /dev/null +++ b/docs/upcoming_changes/8462.improvement.rst @@ -0,0 +1,6 @@ +Updates to ``numba.core.pythonapi``. +==================================== + +Support for Python C-API functions ``PyBytes_AsString`` and +``PyBytes_AsStringAndSize`` is added to ``numba.core.pythonapi.PythonAPI`` as +``bytes_as_string`` and ``bytes_as_string_and_size`` methods respectively. diff --git a/docs/upcoming_changes/8736.np_support.rst b/docs/upcoming_changes/8736.np_support.rst new file mode 100644 index 00000000000..1b1d3b68304 --- /dev/null +++ b/docs/upcoming_changes/8736.np_support.rst @@ -0,0 +1,5 @@ +Support is added for ``numpy.lib.stride_tricks.sliding_window_view``. +===================================================================== + +Support is added for ``numpy.lib.stride_tricks.sliding_window_view`` with +arguments ``window_size`` and ``axis``. diff --git a/docs/upcoming_changes/8885.bug_fix.rst b/docs/upcoming_changes/8885.bug_fix.rst new file mode 100644 index 00000000000..d18244eed3f --- /dev/null +++ b/docs/upcoming_changes/8885.bug_fix.rst @@ -0,0 +1,5 @@ +The ``numpy.allclose`` implementation now correctly handles default arguments. +============================================================================== + +The implementation of ``numpy.allclose`` is corrected to use ``TypingError`` to +report typing errors. diff --git a/docs/upcoming_changes/8892.new_feature.rst b/docs/upcoming_changes/8892.new_feature.rst new file mode 100644 index 00000000000..3546e1f415e --- /dev/null +++ b/docs/upcoming_changes/8892.new_feature.rst @@ -0,0 +1,8 @@ +``numba.experimental.jitclass`` gains support for ``__*matmul__`` methods. +========================================================================== + +``numba.experimental.jitclass`` now has support for the following methods: + +* ``__matmul__`` +* ``__imatmul__`` +* ``__rmatmul__`` diff --git a/docs/upcoming_changes/8906.new_feature.rst b/docs/upcoming_changes/8906.new_feature.rst new file mode 100644 index 00000000000..6ba992b7d60 --- /dev/null +++ b/docs/upcoming_changes/8906.new_feature.rst @@ -0,0 +1,17 @@ +``numba.experimental.jitclass`` gains support for reflected "dunder" methods. +============================================================================= + +``numba.experimental.jitclass`` now has support for the following methods: + +* ``__radd__`` +* ``__rand_`` +* ``__rfloordiv__`` +* ``__rlshift__`` +* ``__ror_`` +* ``__rmod_`` +* ``__rmul_`` +* ``__rpow_`` +* ``__rrshift_`` +* ``__rsub_`` +* ``__rtruediv_`` +* ``__rxor_`` diff --git a/docs/upcoming_changes/8911.improvement.rst b/docs/upcoming_changes/8911.improvement.rst new file mode 100644 index 00000000000..48cdd9f9afb --- /dev/null +++ b/docs/upcoming_changes/8911.improvement.rst @@ -0,0 +1,5 @@ +Support for ``isinstance`` is now non-experimental. +=================================================== + +Support for the ``isinstance`` built-in function has moved from being considered +an experimental feature to a fully supported feature. diff --git a/docs/upcoming_changes/8916.change.rst b/docs/upcoming_changes/8916.change.rst new file mode 100644 index 00000000000..94862a2765e --- /dev/null +++ b/docs/upcoming_changes/8916.change.rst @@ -0,0 +1,4 @@ +The minimum ``llvmlite`` version is now 0.41.0. +=============================================== + +The minimum required version of ``llvmlite`` is now version 0.41.0. diff --git a/docs/upcoming_changes/8944.bug_fix.rst b/docs/upcoming_changes/8944.bug_fix.rst new file mode 100644 index 00000000000..57442f4b3ce --- /dev/null +++ b/docs/upcoming_changes/8944.bug_fix.rst @@ -0,0 +1,4 @@ +Add type validation to ``numpy.isclose``. +========================================= + +Type validation is added to the implementation of ``numpy.isclose``. diff --git a/docs/upcoming_changes/8974.cuda.rst b/docs/upcoming_changes/8974.cuda.rst new file mode 100644 index 00000000000..b4c2cefacbd --- /dev/null +++ b/docs/upcoming_changes/8974.cuda.rst @@ -0,0 +1,13 @@ +Bitwise operation ``ufunc`` support for the CUDA target. +======================================================== + +Support is added for some ``ufunc``s associated with bitwise operation on the +CUDA target. Namely: + +* ``numpy.bitwise_and`` +* ``numpy.bitwise_or`` +* ``numpy.bitwise_not`` +* ``numpy.bitwise_xor`` +* ``numpy.invert`` +* ``numpy.left_shift`` +* ``numpy.right_shift`` diff --git a/docs/upcoming_changes/8988.cuda.rst b/docs/upcoming_changes/8988.cuda.rst new file mode 100644 index 00000000000..eb1990b66f1 --- /dev/null +++ b/docs/upcoming_changes/8988.cuda.rst @@ -0,0 +1,4 @@ +Add support for the latest CUDA driver codes. +============================================= + +Support is added for the latest set of CUDA driver codes. diff --git a/docs/upcoming_changes/8995.np_support.rst b/docs/upcoming_changes/8995.np_support.rst new file mode 100644 index 00000000000..08d016b31b4 --- /dev/null +++ b/docs/upcoming_changes/8995.np_support.rst @@ -0,0 +1,7 @@ +``@vectorize`` accommodates arguments implementing ``__array_ufunc__``. +======================================================================= + +Universal functions (``ufunc``s) created with ``numba.vectorize`` will now +respect arguments implementing ``__array_ufunc__`` (NEP-13) to allow pre- and +post-processing of arguments and return values when the ufunc is called from the +interpreter. diff --git a/docs/upcoming_changes/9035.cuda.rst b/docs/upcoming_changes/9035.cuda.rst new file mode 100644 index 00000000000..5260fb3b3f9 --- /dev/null +++ b/docs/upcoming_changes/9035.cuda.rst @@ -0,0 +1,6 @@ +Add debuginfo support to nvdisasm output. +========================================= + +Support is added for debuginfo (source line and inlining information) in +functions that make calls through ``nvdisasm``. For example the CUDA dispatcher +``.inspect_sass`` method output is now augmented with this information. From 4396a6cbb8962e5d3c4585d19d08824e9920e4c3 Mon Sep 17 00:00:00 2001 From: Siu Kwan Lam <1929845+sklam@users.noreply.github.com> Date: Mon, 31 Jul 2023 14:15:06 -0500 Subject: [PATCH 2/2] Fixups - remove 8736 which is part of the previous release - fix syntax in 8974 and 8995 --- docs/upcoming_changes/8736.np_support.rst | 5 ----- docs/upcoming_changes/8974.cuda.rst | 2 +- docs/upcoming_changes/8995.np_support.rst | 2 +- 3 files changed, 2 insertions(+), 7 deletions(-) delete mode 100644 docs/upcoming_changes/8736.np_support.rst diff --git a/docs/upcoming_changes/8736.np_support.rst b/docs/upcoming_changes/8736.np_support.rst deleted file mode 100644 index 1b1d3b68304..00000000000 --- a/docs/upcoming_changes/8736.np_support.rst +++ /dev/null @@ -1,5 +0,0 @@ -Support is added for ``numpy.lib.stride_tricks.sliding_window_view``. -===================================================================== - -Support is added for ``numpy.lib.stride_tricks.sliding_window_view`` with -arguments ``window_size`` and ``axis``. diff --git a/docs/upcoming_changes/8974.cuda.rst b/docs/upcoming_changes/8974.cuda.rst index b4c2cefacbd..b26b6225bf7 100644 --- a/docs/upcoming_changes/8974.cuda.rst +++ b/docs/upcoming_changes/8974.cuda.rst @@ -1,7 +1,7 @@ Bitwise operation ``ufunc`` support for the CUDA target. ======================================================== -Support is added for some ``ufunc``s associated with bitwise operation on the +Support is added for some ``ufunc``\ s associated with bitwise operation on the CUDA target. Namely: * ``numpy.bitwise_and`` diff --git a/docs/upcoming_changes/8995.np_support.rst b/docs/upcoming_changes/8995.np_support.rst index 08d016b31b4..47a3169108a 100644 --- a/docs/upcoming_changes/8995.np_support.rst +++ b/docs/upcoming_changes/8995.np_support.rst @@ -1,7 +1,7 @@ ``@vectorize`` accommodates arguments implementing ``__array_ufunc__``. ======================================================================= -Universal functions (``ufunc``s) created with ``numba.vectorize`` will now +Universal functions (``ufunc``\ s) created with ``numba.vectorize`` will now respect arguments implementing ``__array_ufunc__`` (NEP-13) to allow pre- and post-processing of arguments and return values when the ufunc is called from the interpreter.