kernelcreation.py 13.9 KB
Newer Older
Martin Bauer's avatar
Martin Bauer committed
1
import itertools
Martin Bauer's avatar
Martin Bauer committed
2
from types import MappingProxyType
Martin Bauer's avatar
Martin Bauer committed
3

4
import sympy as sp
Martin Bauer's avatar
Martin Bauer committed
5

Martin Bauer's avatar
Martin Bauer committed
6
from pystencils.assignment import Assignment
Martin Bauer's avatar
Martin Bauer committed
7
from pystencils.astnodes import Block, Conditional, LoopOverCoordinate, SympyAssignment
Martin Bauer's avatar
Martin Bauer committed
8
from pystencils.cpu.vectorization import vectorize
Martin Bauer's avatar
Martin Bauer committed
9
from pystencils.gpucuda.indexing import indexing_creator_from_params
Martin Bauer's avatar
Martin Bauer committed
10
11
12
from pystencils.simp.assignment_collection import AssignmentCollection
from pystencils.transformations import (
    loop_blocking, move_constants_before_loop, remove_conditionals_in_staggered_kernel)
Martin Bauer's avatar
Martin Bauer committed
13
14


Martin Bauer's avatar
Martin Bauer committed
15
def create_kernel(assignments, target='cpu', data_type="double", iteration_slice=None, ghost_layers=None,
16
                  skip_independence_check=False,
Martin Bauer's avatar
Martin Bauer committed
17
                  cpu_openmp=False, cpu_vectorize_info=None, cpu_blocking=None,
Martin Bauer's avatar
Martin Bauer committed
18
                  gpu_indexing='block', gpu_indexing_params=MappingProxyType({})):
Martin Bauer's avatar
Martin Bauer committed
19
20
    """
    Creates abstract syntax tree (AST) of kernel, using a list of update equations.
21
22

    Args:
Martin Bauer's avatar
Martin Bauer committed
23
        assignments: can be a single assignment, sequence of assignments or an `AssignmentCollection`
24
25
26
27
28
29
30
        target: 'cpu', 'llvm' or 'gpu'
        data_type: data type used for all untyped symbols (i.e. non-fields), can also be a dict from symbol name
                  to type
        iteration_slice: rectangular subset to iterate over, if not specified the complete non-ghost layer \
                         part of the field is iterated over
        ghost_layers: if left to default, the number of necessary ghost layers is determined automatically
                     a single integer specifies the ghost layer count at all borders, can also be a sequence of
Martin Bauer's avatar
Martin Bauer committed
31
                     pairs ``[(x_lower_gl, x_upper_gl), .... ]``
32
33
        skip_independence_check: don't check that loop iterations are independent. This is needed e.g. for
                                 periodicity kernel, that access the field outside the iteration bounds. Use with care!
34
        cpu_openmp: True or number of threads for OpenMP parallelization, False for no OpenMP
Martin Bauer's avatar
Martin Bauer committed
35
36
        cpu_vectorize_info: a dictionary with keys, 'vector_instruction_set', 'assume_aligned' and 'nontemporal'
                            for documentation of these parameters see vectorize function. Example:
37
                            '{'instruction_set': 'avx512', 'assume_aligned': True, 'nontemporal':True}'
Martin Bauer's avatar
Martin Bauer committed
38
        cpu_blocking: a tuple of block sizes or None if no blocking should be applied
Martin Bauer's avatar
Martin Bauer committed
39
        gpu_indexing: either 'block' or 'line' , or custom indexing class, see `AbstractIndexing`
40
        gpu_indexing_params: dict with indexing parameters (constructor parameters of indexing class)
Martin Bauer's avatar
Martin Bauer committed
41
                             e.g. for 'block' one can specify '{'block_size': (20, 20, 10) }'
42
43

    Returns:
Martin Bauer's avatar
Martin Bauer committed
44
        abstract syntax tree (AST) object, that can either be printed as source code with `show_code` or
45
        can be compiled with through its 'compile()' member
Martin Bauer's avatar
Martin Bauer committed
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61

    Example:
        >>> import pystencils as ps
        >>> import numpy as np
        >>> s, d = ps.fields('s, d: [2D]')
        >>> assignment = ps.Assignment(d[0,0], s[0, 1] + s[0, -1] + s[1, 0] + s[-1, 0])
        >>> ast = ps.create_kernel(assignment, target='cpu', cpu_openmp=True)
        >>> kernel = ast.compile()
        >>> d_arr = np.zeros([5, 5])
        >>> kernel(d=d_arr, s=np.ones([5, 5]))
        >>> d_arr
        array([[0., 0., 0., 0., 0.],
               [0., 4., 4., 4., 0.],
               [0., 4., 4., 4., 0.],
               [0., 4., 4., 4., 0.],
               [0., 0., 0., 0., 0.]])
Martin Bauer's avatar
Martin Bauer committed
62
63
    """
    # ----  Normalizing parameters
Martin Bauer's avatar
Martin Bauer committed
64
    split_groups = ()
Martin Bauer's avatar
Martin Bauer committed
65
66
67
68
69
70
    if isinstance(assignments, AssignmentCollection):
        if 'split_groups' in assignments.simplification_hints:
            split_groups = assignments.simplification_hints['split_groups']
        assignments = assignments.all_assignments
    if isinstance(assignments, Assignment):
        assignments = [assignments]
Martin Bauer's avatar
Martin Bauer committed
71
72
73

    # ----  Creating ast
    if target == 'cpu':
Martin Bauer's avatar
Martin Bauer committed
74
75
        from pystencils.cpu import create_kernel
        from pystencils.cpu import add_openmp
Martin Bauer's avatar
Martin Bauer committed
76
        ast = create_kernel(assignments, type_info=data_type, split_groups=split_groups,
77
78
                            iteration_slice=iteration_slice, ghost_layers=ghost_layers,
                            skip_independence_check=skip_independence_check)
Martin Bauer's avatar
Martin Bauer committed
79
80
81
        omp_collapse = None
        if cpu_blocking:
            omp_collapse = loop_blocking(ast, cpu_blocking)
Martin Bauer's avatar
Martin Bauer committed
82
        if cpu_openmp:
Martin Bauer's avatar
Martin Bauer committed
83
            add_openmp(ast, num_threads=cpu_openmp, collapse=omp_collapse)
Martin Bauer's avatar
Martin Bauer committed
84
        if cpu_vectorize_info:
Martin Bauer's avatar
Martin Bauer committed
85
            if cpu_vectorize_info is True:
86
                vectorize(ast)
Martin Bauer's avatar
Martin Bauer committed
87
88
89
90
            elif isinstance(cpu_vectorize_info, dict):
                vectorize(ast, **cpu_vectorize_info)
            else:
                raise ValueError("Invalid value for cpu_vectorize_info")
Martin Bauer's avatar
Martin Bauer committed
91
92
        return ast
    elif target == 'llvm':
Martin Bauer's avatar
Martin Bauer committed
93
        from pystencils.llvm import create_kernel
Martin Bauer's avatar
Martin Bauer committed
94
        ast = create_kernel(assignments, type_info=data_type, split_groups=split_groups,
Martin Bauer's avatar
Martin Bauer committed
95
                            iteration_slice=iteration_slice, ghost_layers=ghost_layers)
Martin Bauer's avatar
Martin Bauer committed
96
97
        return ast
    elif target == 'gpu':
Martin Bauer's avatar
Martin Bauer committed
98
        from pystencils.gpucuda import create_cuda_kernel
Martin Bauer's avatar
Martin Bauer committed
99
        ast = create_cuda_kernel(assignments, type_info=data_type,
Martin Bauer's avatar
Martin Bauer committed
100
                                 indexing_creator=indexing_creator_from_params(gpu_indexing, gpu_indexing_params),
101
102
                                 iteration_slice=iteration_slice, ghost_layers=ghost_layers,
                                 skip_independence_check=skip_independence_check)
Martin Bauer's avatar
Martin Bauer committed
103
104
105
106
107
        return ast
    else:
        raise ValueError("Unknown target %s. Has to be one of 'cpu', 'gpu' or 'llvm' " % (target,))


Martin Bauer's avatar
Martin Bauer committed
108
def create_indexed_kernel(assignments, index_fields, target='cpu', data_type="double", coordinate_names=('x', 'y', 'z'),
Martin Bauer's avatar
Martin Bauer committed
109
                          cpu_openmp=True, gpu_indexing='block', gpu_indexing_params=MappingProxyType({})):
Martin Bauer's avatar
Martin Bauer committed
110
    """
Martin Bauer's avatar
Martin Bauer committed
111
    Similar to :func:`create_kernel`, but here not all cells of a field are updated but only cells with
Martin Bauer's avatar
Martin Bauer committed
112
113
    coordinates which are stored in an index field. This traversal method can e.g. be used for boundary handling.

Martin Bauer's avatar
Martin Bauer committed
114
    The coordinates are stored in a separated index_field, which is a one dimensional array with struct data type.
Martin Bauer's avatar
Martin Bauer committed
115
    This struct has to contain fields named 'x', 'y' and for 3D fields ('z'). These names are configurable with the
Martin Bauer's avatar
Martin Bauer committed
116
    'coordinate_names' parameter. The struct can have also other fields that can be read and written in the kernel, for
Martin Bauer's avatar
Martin Bauer committed
117
118
    example boundary parameters.

Martin Bauer's avatar
Martin Bauer committed
119
120
    index_fields: list of index fields, i.e. 1D fields with struct data type
    coordinate_names: name of the coordinate fields in the struct data type
Martin Bauer's avatar
Martin Bauer committed
121

Martin Bauer's avatar
Martin Bauer committed
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
    Example:
        >>> import pystencils as ps
        >>> import numpy as np
        >>>
        >>> # Index field stores the indices of the cell to visit together with optional values
        >>> index_arr_dtype = np.dtype([('x', np.int32), ('y', np.int32), ('val', np.double)])
        >>> index_arr = np.array([(1, 1, 0.1), (2, 2, 0.2), (3, 3, 0.3)], dtype=index_arr_dtype)
        >>> idx_field = ps.fields(idx=index_arr)
        >>>
        >>> # Additional values  stored in index field can be accessed in the kernel as well
        >>> s, d = ps.fields('s, d: [2D]')
        >>> assignment = ps.Assignment(d[0,0], 2 * s[0, 1] + 2 * s[1, 0] + idx_field('val'))
        >>> ast = create_indexed_kernel(assignment, [idx_field], coordinate_names=('x', 'y'))
        >>> kernel = ast.compile()
        >>> d_arr = np.zeros([5, 5])
        >>> kernel(s=np.ones([5, 5]), d=d_arr, idx=index_arr)
        >>> d_arr
        array([[0. , 0. , 0. , 0. , 0. ],
               [0. , 4.1, 0. , 0. , 0. ],
               [0. , 0. , 4.2, 0. , 0. ],
               [0. , 0. , 0. , 4.3, 0. ],
               [0. , 0. , 0. , 0. , 0. ]])
    """
    if isinstance(assignments, Assignment):
        assignments = [assignments]
    elif isinstance(assignments, AssignmentCollection):
Martin Bauer's avatar
Martin Bauer committed
148
        assignments = assignments.all_assignments
Martin Bauer's avatar
Martin Bauer committed
149
    if target == 'cpu':
Martin Bauer's avatar
Martin Bauer committed
150
151
152
153
154
155
        from pystencils.cpu import create_indexed_kernel
        from pystencils.cpu import add_openmp
        ast = create_indexed_kernel(assignments, index_fields=index_fields, type_info=data_type,
                                    coordinate_names=coordinate_names)
        if cpu_openmp:
            add_openmp(ast, num_threads=cpu_openmp)
Martin Bauer's avatar
Martin Bauer committed
156
157
158
159
        return ast
    elif target == 'llvm':
        raise NotImplementedError("Indexed kernels are not yet supported in LLVM backend")
    elif target == 'gpu':
Martin Bauer's avatar
Martin Bauer committed
160
        from pystencils.gpucuda import created_indexed_cuda_kernel
Martin Bauer's avatar
Martin Bauer committed
161
162
163
        idx_creator = indexing_creator_from_params(gpu_indexing, gpu_indexing_params)
        ast = created_indexed_cuda_kernel(assignments, index_fields, type_info=data_type,
                                          coordinate_names=coordinate_names, indexing_creator=idx_creator)
Martin Bauer's avatar
Martin Bauer committed
164
165
166
        return ast
    else:
        raise ValueError("Unknown target %s. Has to be either 'cpu' or 'gpu'" % (target,))
167

168

169
170
def create_staggered_kernel(staggered_field, expressions, subexpressions=(), target='cpu',
                            gpu_exclusive_conditions=False, **kwargs):
171
172
    """Kernel that updates a staggered field.

Martin Bauer's avatar
Martin Bauer committed
173
174
    .. image:: /img/staggered_grid.svg

175
    Args:
176
        staggered_field: field where the first index coordinate defines the location of the staggered value
177
178
                can have 1 or 2 index coordinates, in case of two index coordinates at every staggered location
                a vector is stored, expressions parameter has to be a sequence of sequences then
Martin Bauer's avatar
Martin Bauer committed
179
180
                where e.g. ``f[0,0](0)`` is interpreted as value at the left cell boundary, ``f[1,0](0)`` the right cell
                boundary and ``f[0,0](1)`` the southern cell boundary etc.
181
        expressions: sequence of expressions of length dim, defining how the west, southern, (bottom) cell boundary
182
                     should be updated.
183
184
        subexpressions: optional sequence of Assignments, that define subexpressions used in the main expressions
        target: 'cpu' or 'gpu'
185
        gpu_exclusive_conditions: if/else construct to have only one code block for each of 2**dim code paths
186
        kwargs: passed directly to create_kernel, iteration slice and ghost_layers parameters are not allowed
187

188
    Returns:
189
        AST, see `create_kernel`
190
191
    """
    assert 'iteration_slice' not in kwargs and 'ghost_layers' not in kwargs
192
    assert staggered_field.index_dimensions in (1, 2), 'Staggered field must have one or two index dimensions'
193
194
195
196
197
    dim = staggered_field.spatial_dimensions

    counters = [LoopOverCoordinate.get_loop_counter_symbol(i) for i in range(dim)]
    conditions = [counters[i] < staggered_field.shape[i] - 1 for i in range(dim)]
    assert len(expressions) == dim
198
199
200
201
202
    if staggered_field.index_dimensions == 2:
        assert all(len(sublist) == len(expressions[0]) for sublist in expressions), \
            "If staggered field has two index dimensions expressions has to be a sequence of sequences of all the " \
            "same length."

203
    final_assignments = []
204
205
206
207
    last_conditional = None

    def add(condition, dimensions, as_else_block=False):
        nonlocal last_conditional
208
        if staggered_field.index_dimensions == 1:
209
210
211
            assignments = [Assignment(staggered_field(d), expressions[d]) for d in dimensions]
            a_coll = AssignmentCollection(assignments, list(subexpressions))
            a_coll = a_coll.new_filtered([staggered_field(d) for d in dimensions])
212
213
        elif staggered_field.index_dimensions == 2:
            assert staggered_field.has_fixed_index_shape
214
215
216
            assignments = [Assignment(staggered_field(d, i), expr)
                           for d in dimensions
                           for i, expr in enumerate(expressions[d])]
217
            a_coll = AssignmentCollection(assignments, list(subexpressions))
218
219
            a_coll = a_coll.new_filtered([staggered_field(d, i) for i in range(staggered_field.index_shape[1])
                                          for d in dimensions])
220
        sp_assignments = [SympyAssignment(a.lhs, a.rhs) for a in a_coll.all_assignments]
221
        if as_else_block and last_conditional:
222
223
224
            new_cond = Conditional(condition, Block(sp_assignments))
            last_conditional.false_block = Block([new_cond])
            last_conditional = new_cond
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
        else:
            last_conditional = Conditional(condition, Block(sp_assignments))
            final_assignments.append(last_conditional)

    if target == 'cpu' or not gpu_exclusive_conditions:
        for d in range(dim):
            cond = sp.And(*[conditions[i] for i in range(dim) if d != i])
            add(cond, [d])
    elif target == 'gpu':
        full_conditions = [sp.And(*[conditions[i] for i in range(dim) if d != i]) for d in range(dim)]
        for include in itertools.product(*[[1, 0]] * dim):
            case_conditions = sp.And(*[c if value else sp.Not(c) for c, value in zip(full_conditions, include)])
            dimensions_to_include = [i for i in range(dim) if include[i]]
            if dimensions_to_include:
                add(case_conditions, dimensions_to_include, True)
240

241
242
    ghost_layers = [(1, 0)] * dim

Martin Bauer's avatar
Martin Bauer committed
243
244
245
246
    blocking = kwargs.get('cpu_blocking', None)
    if blocking:
        del kwargs['cpu_blocking']

247
248
249
    cpu_vectorize_info = kwargs.get('cpu_vectorize_info', None)
    if cpu_vectorize_info:
        del kwargs['cpu_vectorize_info']
250
251
252
253
    openmp = kwargs.get('cpu_openmp', None)
    if openmp:
        del kwargs['cpu_openmp']

254
    ast = create_kernel(final_assignments, ghost_layers=ghost_layers, target=target, **kwargs)
255

256
257
    if target == 'cpu':
        remove_conditionals_in_staggered_kernel(ast)
258
        move_constants_before_loop(ast)
259
        omp_collapse = None
Martin Bauer's avatar
Martin Bauer committed
260
        if blocking:
261
262
263
264
            omp_collapse = loop_blocking(ast, blocking)
        if openmp:
            from pystencils.cpu import add_openmp
            add_openmp(ast, num_threads=openmp, collapse=omp_collapse, assume_single_outer_loop=False)
265
266
267
268
        if cpu_vectorize_info is True:
            vectorize(ast)
        elif isinstance(cpu_vectorize_info, dict):
            vectorize(ast, **cpu_vectorize_info)
269
    return ast