Dpctl

This is a bug-fix release, which also provides a change needed by ``numba_dpex`` project to support dispatching kernels
consuming instances of ``sycl::local_accessor`` template type.

Changed

* Changed behavior of ``dpctl.tensor.usm_ndarray.__dlpack_device__`` method to return device id of the parent unpartitioned device if array is allocated on a sub-device instead of raising an exception: [1604](https://github.com/IntelPython/dpctl/pull/1604)
* Array creation functions and the ``usm_ndarray`` constructor in `dpctl.tensor` submodule now use cached default-selected device to improve performance: [1606](https://github.com/IntelPython/dpctl/pull/1606)
* Changed treatment of `axis` keyword for `dpctl.tensor.tensordot` and `dpctl.tensor.vecdot` to align with Python Array API 2023.12 specification: [1608](https://github.com/IntelPython/dpctl/pull/1608)
* Changed implementation of `DPCTLQueue_SubmitRange`, `DPCTLQueue_SubmitNDRange` in DPCTLSyclInterface library to support ``sycl::local_accessor`` arguments needed by ``numba_dpex``; the enum `DPCTLKernelArgType` to correspond to C++ disjoint types: [1609](https://github.com/IntelPython/dpctl/pull/1609), [#1611](https://github.com/IntelPython/dpctl/pull/1611), [#1612](https://github.com/IntelPython/dpctl/pull/1612)

Fixed

* Fixed a crash on Windows platform during execution of getter of `dpctl.SyclPlatfom.default_context` property: : [1604](https://github.com/IntelPython/dpctl/pull/1604)
* Fixed kernel submission error on NVidia CUDA GPUs during `dpctl.tensor.matmul` operation: [1605](https://github.com/IntelPython/dpctl/pull/1605)
* Fixed corruption of context cache table entries: [1607](https://github.com/IntelPython/dpctl/pull/1607)
* Fixed incorrect result from ``dpctl.tensor.tensordot`` reported in issue [1570](https://github.com/IntelPython/dpctl/issues/1570): [#1608](https://github.com/IntelPython/dpctl/pull/1608)
* Fixed library name output by ``python -m dpctl --library``: [1615](https://github.com/IntelPython/dpctl/pull/1615)

This release will require DPC++ 2024.1.0, which no longer supports Intel Gen9 integrated GPUs found in Intel CPUs of 10th generation and older.
Featurewise, this release is identical to 0.15.1.

This release reaches milestone of 100% compliance of `dpctl.tensor` functions with Python Array API 2022.12 standard for the main namespace.

Added

* Added reduction functions `dpctl.tensor.min`, `dpctl.tensor.max`, `dpctl.tensor.argmin`, `dpctl.tensor.argmax`, and `dpctl.tensor.prod` per Python Array API specifications: [1399](https://github.com/IntelPython/dpctl/pull/1399)
* Added dedicated in-place operations for binary elementwise operations and deployed them in Python operators of `dpctl.tensor.usm_ndarray` type: [1431](https://github.com/IntelPython/dpctl/pull/1431), [#1447](https://github.com/IntelPython/dpctl/pull/1447)
* Added new elementwise functions `dpctl.tensor.cbrt`, `dpctl.tensor.rsqrt`, `dpctl.tensor.exp2`, `dpctl.tensor.copysign`, `dpctl.tensor.angle`, and `dpctl.tensor.reciprocal`: [1443](https://github.com/IntelPython/dpctl/pull/1443), [#1474](https://github.com/IntelPython/dpctl/pull/1474)
* Added statistical functions `dpctl.tensor.mean`, `dpctl.tensor.std`, `dpctl.tensor.var` per Python Array API specifications: [1465](https://github.com/IntelPython/dpctl/pull/1465)
* Added sorting functions `dpctl.tensor.sort` and `dpctl.tensor.argsort`, and set functions `dpctl.tensor.unique_values`, `dpctl.tensor.unique_counts`, `dpctl.tensor.unique_inverse`, `dpctl.tensor.unique_all`: [1483](https://github.com/IntelPython/dpctl/pull/1483)
* Added linear algebra functions from the Array API namespace `dpctl.tensor.matrix_transpose`, `dpctl.tensor.matmul`, `dpctl.tensor.vecdot`, and `dpctl.tensor.tensordot`: [1490](https://github.com/IntelPython/dpctl/pull/1490), [#1525](https://github.com/IntelPython/dpctl/pull/1525), [#1541](https://github.com/IntelPython/dpctl/pull/1541)
* Added `dpctl.tensor.clip` function: [1444](https://github.com/IntelPython/dpctl/pull/1444), [#1505](https://github.com/IntelPython/dpctl/pull/1505)
* Added custom reduction functions `dpt.logsumexp` (reduction using binary function `dpctl.tensor.logaddexp`), `dpt.reduce_hypot` (reduction using binary function `dpctl.tensor.hypot`): [1446](https://github.com/IntelPython/dpctl/pull/1446)
* Added inspection API to query capabilities of Python Array API specification implementation: [1469](https://github.com/IntelPython/dpctl/pull/1469)
* Support for compilation for NVIDIA(R) sycl target with use of [CodePlay oneAPI plug-in](https://developer.codeplay.com/products/oneapi/nvidia/home/): [#1411](https://github.com/IntelPython/dpctl/pull/1411), [#1124](https://github.com/IntelPython/dpctl/discussions/1124)
* Added `dpctl.utils.intel_device_info` function to query additional information about Intel(R) GPU devices: [gh-1428](https://github.com/IntelPython/dpctl/pull/1428) and [gh-1445](https://github.com/IntelPython/dpctl/pull/1445)
* Added support for two new device descriptors, `dpctl.SyclDevice.max_mem_alloc_size` and `dpctl.SyclDevice.max_clock_frequency`: [1530](https://github.com/IntelPython/dpctl/pull/1530)

Changed

* Functions `dpctl.tensor.result_type` and `dpctl.tensor.can_cast` became device-aware: [1488](https://github.com/IntelPython/dpctl/pull/1488), [#1473](https://github.com/IntelPython/dpctl/pull/1473)
* Implementation of method `dpctl.SyclEvent.wait_for` changed to use ``sycl::event::wait`` instead of ``sycl::event::wait_and_throw``: [gh-1436](https://github.com/IntelPython/dpctl/pull/1436)
* `dpctl.tensor.astype` was changed to support `device` keyword as per Python Array API specification: [1511](https://github.com/IntelPython/dpctl/pull/1511)
* C++ header files in `libtensor/include/kernels` containing implementations of SYCL kernels no longer depends on "pybind11.h": [1516](https://github.com/IntelPython/dpctl/pull/1516)

Fixed

* Fixed issues with `dpctl.tensor.repeat` support for `axis` keyword: [1427](https://github.com/IntelPython/dpctl/pull/1427), [#1433](https://github.com/IntelPython/dpctl/pull/1433)
* Fix for gh-1503 for bug `usm_ndarray.__setitem__`: [1504](https://github.com/IntelPython/dpctl/pull/1504)
* Other bug fixes: [1485](https://github.com/IntelPython/dpctl/pull/1485), [#1477](https://github.com/IntelPython/dpctl/pull/1477), [#1512](https://github.com/IntelPython/dpctl/pull/1512)

Added

* Added `dpctl.tensor.floor`, `dpctl.tensor.ceil`, `dpctl.tensor.trunc` elementwise functions.
* Added `dpctl.tensor.hypot`, `dpctl.tensor.logaddexp` elementwise functions.
* Added trigonometric (`dpctl.tensor.sin`, `dpctl.tensor.cos`, `dpctl.tensor.tan`) and hyperbolic (`dpctl.tensor.sinh`, `dpctl.tensor.cosh`, `dpctl.tensor.tanh`) elementwise functions and their inverses (`dpctl.tensor.asin`, `dpctl.tensor.asinh`, `dpctl.tensor.acos`, `dpctl.tensor.acosh`, `dpctl.tensor.atan`, `dpctl.tensor.atanh`).
* Added `dpctl.tensor.round` function.
* Added `dpctl.tensor.sign` and `dpctl.tensor.remainder` elementwise functions.
* Added bitwise elementwise functions `dpctl.tensor.bitwise_and`, `dpctl.tensor.bitwise_xor`, `dpctl.tensor.bitwise_or`, `dpctl.tensor.bitwise_invert`
* Added bitwise shift functions `dpctl.tensor.bitwise_left_shift` and `dpctl.tensor.bitwise_right_shift`.
* Added `dpctl.tensor.atan2` and `dpctl.tensor.signbit` elementwise functions.
* Added `dpctl.tensor.minumum` and `dpctl.tensor.maximum` binary elementwise functions.
* Supported equality checking and hashing for `dpctl.SyclPlatform`.
* Implemented `types` property for all unary and binary elementwise functions [1361](https://github.com/IntelPython/dpctl/pull/1361)
* Added `dpctl.tensor.repeat` and `dpctl.tensor.tile` functions.
* Added `dpctl.tensor.matrix_transpose ` function.

Changed

* Enabled support for Python arithmetic, in-place arithmetic, reflexive arithmetic, comparison, and bitwise operators for `dpctl.tensor.usm_ndarray` type [1324](https://github.com/IntelPython/dpctl/pull/1324).
* Removed `dpctl.tensor.numpy_usm_shared` obsolete class and associated tests which were being skipped [1310](https://github.com/IntelPython/dpctl/pull/1310)
* Transitioned `dpctl` codebase to Cython 3.
* Improved performance of boolean reduction functions `dpctl.tensor.all` and `dpctl.tensor.any`.
* Improved performance of summation function `dpctl.tensor.sum`.
* Improved in-place arithmetic operations for addition, subtraction and multiplication.
* Updated codebase per SYCL-2020 intel/llvm compiler deprecation warnings.
* Improved performance of advanced boolean indexing for arrays whose size fits in 32-bit signed integer type.
* Removed deprecated `DPCTLDevice_GetMaxWorkItemSizes` function from the SyclInterface library.
* Improved performance of `dpctl.tensor.reshape` in the case when a copy is being made.
* Improved performance of `dpctl.tensor.roll` function.

Fixed

* Fixed issues identified by Coverity security scans.
* Fixed issues [1279](https://github.com/IntelPython/dpctl/issues/1279), [#1350](https://github.com/IntelPython/dpctl/issues/1350), [#1344](https://github.com/IntelPython/dpctl/issues/1344), [#1327](https://github.com/IntelPython/dpctl/issues/1327), [#1241](https://github.com/IntelPython/dpctl/issues/1241), [#1250](https://github.com/IntelPython/dpctl/issues/1250), [#1293](https://github.com/IntelPython/dpctl/issues/1293).

Added

* Added `dpctl.tensor.log2` and `dpctl.tensor.log10`: [1267](https://github.com/IntelPython/dpctl/pull/1267)
* Added `dpctl.tensor.negative`, `dpctl.tensor.positive`, `dpctl.tensor.square` [1268](https://github.com/IntelPython/dpctl/pull/1268)
* Added `dpctl.tensor.logical_not`, `dpctl.tensor.logical_and`, `dpctl.tensor.logical_or`, `dpctl.tensor.logical_xor` [1270](https://github.com/IntelPython/dpctl/pull/1270)

Changed

* `dpctl.tensor.astype` behavior for `newdtype=None` changes [1261](https://github.com/IntelPython/dpctl/pull/1262)
* `dpctl.tensor.usm_ndaray` constructor default value of `dtype` keyword argument changed to `None`: [1265](https://github.com/IntelPython/dpctl/pull/1265)
* Support for `out` arguments that overlap with inputs for unary elementwise functions[1281](https://github.com/IntelPython/dpctl/pull/1281)
* Copying from one array to another a no-op if both arrays view into the same memory [1284](https://github.com/IntelPython/dpctl/pull/1284)

Added

* Added `dpctl.tensor.less_equal`, `dpctl.tensor.greater`, `dpctl.tensor.greater_equal`: [1239](https://github.com/IntelPython/dpctl/pull/1239)

Changed

* Optimized in-place arithmetic operations for updating matrix with rows/columns via broadcasting: [1244](https://github.com/IntelPython/dpctl/pull/1244)

Fixed

* Fixed handling of 0d arrays in `dpctl.tensor.sum`: [1238](https://github.com/IntelPython/dpctl/pull/1238)