4880232227

Build Type

github

Committed by GitHub

Commit Message

Merge b5b70b35c into 20d4340f6

Pull Request Pull Request #1029: Major housekeeping

Run Details

1548 of 3274 branches covered (47.28%)

Branch coverage included in aggregate %.

1 of 1 new or added line in 1 file covered. (100.0%)

666 existing lines in 13 files now uncovered.

5399 of 8971 relevant lines covered (60.18%)

0.6 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

73.02

/numba_dpex/core/kernel_interface/utils.py

# SPDX-FileCopyrightText: 2022 - 2023 Intel Corporation
#
# SPDX-License-Identifier: Apache-2.0

from warnings import warn

import dpctl
from numba.core.types import Array as NpArrayType

from numba_dpex.core.exceptions import (
    ComputeFollowsDataInferenceError,
    ExecutionQueueInferenceError,
)
from numba_dpex.core.types import USMNdArray


def chk_compute_follows_data_compliance(usm_array_arglist):
    """Check if all the usm ndarray's have the same device.

    Extracts the device filter string from the Numba inferred USMNdArray
    type. Check if the devices corresponding to the filter string are
    equivalent and return a ``dpctl.SyclDevice`` object corresponding to the
    common filter string.

    If an exception occurred in creating a ``dpctl.SyclDevice``, or the
    devices are not equivalent then returns None.

    Args:
        usm_array_arglist : A list of usm_ndarray types specified as
        arguments to the kernel.

    Returns:
        A ``dpctl.SyclDevice`` object if all USMNdArray have same device, or
        else None is returned.
    """

    queue = None

    for usm_array in usm_array_arglist:
        _queue = usm_array.queue
        if not queue:
            queue = _queue
        else:
            if _queue != queue:
                return None

    return queue


def determine_kernel_launch_queue(args, argtypes, kernel_name):
    """Determines the queue where the kernel is to be launched.

    The execution queue is derived using the following algorithm. In future,
    support for ``numpy.ndarray`` and ``dpctl.device_context`` is to be
    removed and queue derivation will follows Python Array API's
    "compute follows data" logic.

    Check if there are array arguments.
    True:
        Check if all array arguments are of type numpy.ndarray
        (numba.types.Array)
            True:
                Check if the kernel was invoked from within a
                dpctl.device_context.
                True:
                    Provide a deprecation warning for device_context use and
                    point to using dpctl.tensor.usm_ndarray or dpnp.ndarray

                    return dpctl.get_current_queue
                False:
                    Raise ExecutionQueueInferenceError
            False:
                Check if all of the arrays are USMNdarray
                    True:
                        Check if execution queue could be inferred using
                        compute follows data rules
                        True:
                            return the compute follows data inferred queue
                        False:
                            Raise ComputeFollowsDataInferenceError
                    False:
                        Raise ComputeFollowsDataInferenceError
    False:
        Check if the kernel was invoked from within a dpctl.device_context.
        True:
            Provide a deprecation warning for device_context use and
            point to using dpctl.tensor.usm_ndarray of dpnp.ndarray

            return dpctl.get_current_queue
        False:
            Raise ExecutionQueueInferenceError

    Args:
        args : A list of arguments passed to the kernel stored in the
        launcher.
        argtypes : The Numba inferred type for each argument.

    Returns:
        A queue the common queue used to allocate the arrays. If no such
        queue exists, then raises an Exception.

    Raises:
        ComputeFollowsDataInferenceError: If the queue could not be inferred
            using compute follows data rules.
        ExecutionQueueInferenceError: If the queue could not be inferred
            using the dpctl queue manager.
    """

    # FIXME: The args parameter is not needed once numpy support is removed

    # Needed as USMNdArray derives from Array
    array_argnums = [
        i
        for i, _ in enumerate(args)
        if isinstance(argtypes[i], NpArrayType)
        and not isinstance(argtypes[i], USMNdArray)
    ]
    usmarray_argnums = [
        i for i, _ in enumerate(args) if isinstance(argtypes[i], USMNdArray)
    ]

    # if usm and non-usm array arguments are getting mixed, then the
    # execution queue cannot be inferred using compute follows data rules.
    if array_argnums and usmarray_argnums:
        raise ComputeFollowsDataInferenceError(
            array_argnums, usmarray_argnum_list=usmarray_argnums
        )
    elif array_argnums and not usmarray_argnums:
        if dpctl.is_in_device_context():
            warn(
                "Support for dpctl.device_context to specify the "
                + "execution queue is deprecated. "
                + "Use dpctl.tensor.usm_ndarray based array "
                + "containers instead. ",
                DeprecationWarning,
                stacklevel=2,
            )
            warn(
                "Support for NumPy ndarray objects as kernel arguments is "
                + "deprecated. Use dpctl.tensor.usm_ndarray based array "
                + "containers instead. ",
                DeprecationWarning,
                stacklevel=2,
            )
            return dpctl.get_current_queue()
        else:
            raise ExecutionQueueInferenceError(kernel_name)
    elif usmarray_argnums and not array_argnums:
        if dpctl.is_in_device_context():
            warn(
                "dpctl.device_context ignored as the kernel arguments "
                + "are dpctl.tensor.usm_ndarray based array containers."
            )
        usm_array_args = [
            argtype
            for i, argtype in enumerate(argtypes)
            if i in usmarray_argnums
        ]

        queue = chk_compute_follows_data_compliance(usm_array_args)

        if not queue:
            raise ComputeFollowsDataInferenceError(
                kernel_name, usmarray_argnum_list=usmarray_argnums
            )

        return queue
    else:
        if dpctl.is_in_device_context():
            warn(
                "Support for dpctl.device_context to specify the "
                + "execution queue is deprecated. "
                + "Use dpctl.tensor.usm_ndarray based array "
                + "containers instead. ",
                DeprecationWarning,
                stacklevel=2,
            )
            return dpctl.get_current_queue()
        else:
            raise ExecutionQueueInferenceError(kernel_name)

1	# SPDX-FileCopyrightText: 2022 - 2023 Intel Corporation
2	#
3	# SPDX-License-Identifier: Apache-2.0
4
5	from warnings import warn	1✔
6
7	import dpctl	1✔
8	from numba.core.types import Array as NpArrayType	1✔
9
10	from numba_dpex.core.exceptions import (	1✔
11	ComputeFollowsDataInferenceError,
12	ExecutionQueueInferenceError,
13	)
14	from numba_dpex.core.types import USMNdArray	1✔
15
16
17	def chk_compute_follows_data_compliance(usm_array_arglist):	1✔
18	"""Check if all the usm ndarray's have the same device.
19
20	Extracts the device filter string from the Numba inferred USMNdArray
21	type. Check if the devices corresponding to the filter string are
22	equivalent and return a ``dpctl.SyclDevice`` object corresponding to the
23	common filter string.
24
25	If an exception occurred in creating a ``dpctl.SyclDevice``, or the
26	devices are not equivalent then returns None.
27
28	Args:
29	usm_array_arglist : A list of usm_ndarray types specified as
30	arguments to the kernel.
31
32	Returns:
33	A ``dpctl.SyclDevice`` object if all USMNdArray have same device, or
34	else None is returned.
35	"""
36
37	queue = None	1✔
38
39	for usm_array in usm_array_arglist:	1✔
40	_queue = usm_array.queue	1✔
41	if not queue:	1✔
42	queue = _queue	1✔
43	else:
44	if _queue != queue:	1!
45	return None	×
46
47	return queue	1✔
48
49
50	def determine_kernel_launch_queue(args, argtypes, kernel_name):	1✔
51	"""Determines the queue where the kernel is to be launched.
52
53	The execution queue is derived using the following algorithm. In future,
54	support for ``numpy.ndarray`` and ``dpctl.device_context`` is to be
55	removed and queue derivation will follows Python Array API's
56	"compute follows data" logic.
57
58	Check if there are array arguments.
59	True:
60	Check if all array arguments are of type numpy.ndarray
61	(numba.types.Array)
62	True:
63	Check if the kernel was invoked from within a
64	dpctl.device_context.
65	True:
66	Provide a deprecation warning for device_context use and
67	point to using dpctl.tensor.usm_ndarray or dpnp.ndarray
68
69	return dpctl.get_current_queue
70	False:
71	Raise ExecutionQueueInferenceError
72	False:
73	Check if all of the arrays are USMNdarray
74	True:
75	Check if execution queue could be inferred using
76	compute follows data rules
77	True:
78	return the compute follows data inferred queue
79	False:
80	Raise ComputeFollowsDataInferenceError
81	False:
82	Raise ComputeFollowsDataInferenceError
83	False:
84	Check if the kernel was invoked from within a dpctl.device_context.
85	True:
86	Provide a deprecation warning for device_context use and
87	point to using dpctl.tensor.usm_ndarray of dpnp.ndarray
88
89	return dpctl.get_current_queue
90	False:
91	Raise ExecutionQueueInferenceError
92
93	Args:
94	args : A list of arguments passed to the kernel stored in the
95	launcher.
96	argtypes : The Numba inferred type for each argument.
97
98	Returns:
99	A queue the common queue used to allocate the arrays. If no such
100	queue exists, then raises an Exception.
101
102	Raises:
103	ComputeFollowsDataInferenceError: If the queue could not be inferred
104	using compute follows data rules.
105	ExecutionQueueInferenceError: If the queue could not be inferred
106	using the dpctl queue manager.
107	"""
108
109	# FIXME: The args parameter is not needed once numpy support is removed
110
111	# Needed as USMNdArray derives from Array
112	array_argnums = [	1✔
113	i
114	for i, _ in enumerate(args)
115	if isinstance(argtypes[i], NpArrayType)
116	and not isinstance(argtypes[i], USMNdArray)
117	]
118	usmarray_argnums = [	1✔
119	i for i, _ in enumerate(args) if isinstance(argtypes[i], USMNdArray)
120	]
121
122	# if usm and non-usm array arguments are getting mixed, then the
123	# execution queue cannot be inferred using compute follows data rules.
124	if array_argnums and usmarray_argnums:	1!
UNCOV 125	raise ComputeFollowsDataInferenceError(	×
126	array_argnums, usmarray_argnum_list=usmarray_argnums
127	)
128	elif array_argnums and not usmarray_argnums:	1✔
129	if dpctl.is_in_device_context():	1!
130	warn(	1✔
131	"Support for dpctl.device_context to specify the "
132	+ "execution queue is deprecated. "
133	+ "Use dpctl.tensor.usm_ndarray based array "
134	+ "containers instead. ",
135	DeprecationWarning,
136	stacklevel=2,
137	)
138	warn(	1✔
139	"Support for NumPy ndarray objects as kernel arguments is "
140	+ "deprecated. Use dpctl.tensor.usm_ndarray based array "
141	+ "containers instead. ",
142	DeprecationWarning,
143	stacklevel=2,
144	)
145	return dpctl.get_current_queue()	1✔
146	else:
147	raise ExecutionQueueInferenceError(kernel_name)	×
148	elif usmarray_argnums and not array_argnums:	1!
149	if dpctl.is_in_device_context():	1!
UNCOV 150	warn(	×
151	"dpctl.device_context ignored as the kernel arguments "
152	+ "are dpctl.tensor.usm_ndarray based array containers."
153	)
154	usm_array_args = [	1✔
155	argtype
156	for i, argtype in enumerate(argtypes)
157	if i in usmarray_argnums
158	]
159
160	queue = chk_compute_follows_data_compliance(usm_array_args)	1✔
161
162	if not queue:	1!
163	raise ComputeFollowsDataInferenceError(	×
164	kernel_name, usmarray_argnum_list=usmarray_argnums
165	)
166
167	return queue	1✔
168	else:
169	if dpctl.is_in_device_context():	×
170	warn(	×
171	"Support for dpctl.device_context to specify the "
172	+ "execution queue is deprecated. "
173	+ "Use dpctl.tensor.usm_ndarray based array "
174	+ "containers instead. ",
175	DeprecationWarning,
176	stacklevel=2,
177	)
178	return dpctl.get_current_queue()	×
179	else:
180	raise ExecutionQueueInferenceError(kernel_name)	×

IntelPython / numba-dpex / 4880232227

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous