def created_indexed_cuda_kernel(assignments,
index_fields,
function_name="kernel",
type_info=None,
coordinate_names=('x', 'y', 'z'),
indexing_creator=BlockIndexing):
fields_read, fields_written, assignments = add_types(assignments, type_info, check_independence_condition=False)
all_fields = fields_read.union(fields_written)
read_only_fields = set([f.name for f in fields_read - fields_written])
for index_field in index_fields:
index_field.field_type = FieldType.INDEXED
assert FieldType.is_indexed(index_field)
assert index_field.spatial_dimensions == 1, "Index fields have to be 1D"
non_index_fields = [f for f in all_fields if f not in index_fields]
spatial_coordinates = {f.spatial_dimensions for f in non_index_fields}
assert len(spatial_coordinates) == 1, "Non-index fields do not have the same number of spatial coordinates"
spatial_coordinates = list(spatial_coordinates)[0]
def get_coordinate_symbol_assignment(name):
for ind_f in index_fields:
assert isinstance(ind_f.dtype, StructType), "Index fields have to have a struct data type"
data_type = ind_f.dtype
if data_type.has_element(name):
rhs = ind_f[0](name)
lhs = TypedSymbol(name, np.int64)
return SympyAssignment(lhs, rhs)
raise ValueError(f"Index {name} not found in any of the passed index fields")
coordinate_symbol_assignments = [get_coordinate_symbol_assignment(n)
for n in coordinate_names[:spatial_coordinates]]
coordinate_typed_symbols = [eq.lhs for eq in coordinate_symbol_assignments]
idx_field = list(index_fields)[0]
indexing = indexing_creator(field=idx_field,
iteration_slice=[slice(None, None, None)] * len(idx_field.spatial_shape))
function_body = Block(coordinate_symbol_assignments + assignments)
function_body = indexing.guard(function_body, get_common_shape(index_fields))
ast = KernelFunction(function_body, Target.GPU, Backend.CUDA, make_python_function,
None, function_name, assignments=assignments)
ast.global_variables.update(indexing.index_variables)
coord_mapping = indexing.coordinates
base_pointer_spec = [['spatialInner0']]
base_pointer_info = {f.name: parse_base_pointer_info(base_pointer_spec, [2, 1, 0],
f.spatial_dimensions, f.index_dimensions)
for f in all_fields}
coord_mapping = {f.name: coord_mapping for f in index_fields}
coord_mapping.update({f.name: coordinate_typed_symbols for f in non_index_fields})
resolve_field_accesses(ast, read_only_fields, field_to_fixed_coordinates=coord_mapping,
field_to_base_pointer_info=base_pointer_info)
# add the function which determines #blocks and #threads as additional member to KernelFunction node
# this is used by the jit
ast.indexing = indexing
return ast
def create_kernel(assignments: AssignmentOrAstNodeList,
function_name: str = "kernel",
type_info='double',
split_groups=(),
iteration_slice=None,
ghost_layers=None,
skip_independence_check=False) -> KernelFunction:
"""Creates an abstract syntax tree for a kernel function, by taking a list of update rules.
Loops are created according to the field accesses in the equations.
Args:
assignments: list of sympy equations, containing accesses to :class:`pystencils.field.Field`.
Defining the update rules of the kernel
function_name: name of the generated function - only important if generated code is written out
type_info: a map from symbol name to a C type specifier. If not specified all symbols are assumed to
be of type 'double' except symbols which occur on the left hand side of equations where the
right hand side is a sympy Boolean which are assumed to be 'bool' .
split_groups: Specification on how to split up inner loop into multiple loops. For details see
transformation :func:`pystencils.transformation.split_inner_loop`
iteration_slice: if not None, iteration is done only over this slice of the field
ghost_layers: a sequence of pairs for each coordinate with lower and upper nr of ghost layers
if None, the number of ghost layers is determined automatically and assumed to be equal for a
all dimensions
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!
Returns:
AST node representing a function, that can be printed as C or CUDA code
"""
def type_symbol(term):
if isinstance(term, Field.Access) or isinstance(term, TypedSymbol):
return term
elif isinstance(term, sp.Symbol):
if not hasattr(type_info, '__getitem__'):
return TypedSymbol(term.name, create_type(type_info))
else:
return TypedSymbol(term.name, type_info[term.name])
else:
raise ValueError("Term has to be field access or symbol")
fields_read, fields_written, assignments = add_types(
assignments, type_info, not skip_independence_check)
all_fields = fields_read.union(fields_written)
read_only_fields = set([f.name for f in fields_read - fields_written])
buffers = set([f for f in all_fields if FieldType.is_buffer(f)])
fields_without_buffers = all_fields - buffers
body = ast.Block(assignments)
loop_order = get_optimal_loop_ordering(fields_without_buffers)
loop_node, ghost_layer_info = make_loop_over_domain(
body,
iteration_slice=iteration_slice,
ghost_layers=ghost_layers,
loop_order=loop_order)
ast_node = KernelFunction(loop_node,
'cpu',
'c',
compile_function=make_python_function,
ghost_layers=ghost_layer_info,
function_name=function_name)
if split_groups:
typed_split_groups = [[type_symbol(s) for s in split_group]
for split_group in split_groups]
split_inner_loop(ast_node, typed_split_groups)
base_pointer_spec = [['spatialInner0'], ['spatialInner1']
] if len(loop_order) >= 2 else [['spatialInner0']]
base_pointer_info = {
field.name: parse_base_pointer_info(base_pointer_spec, loop_order,
field.spatial_dimensions,
field.index_dimensions)
for field in fields_without_buffers
}
buffer_base_pointer_info = {
field.name: parse_base_pointer_info([['spatialInner0']], [0],
field.spatial_dimensions,
field.index_dimensions)
for field in buffers
}
base_pointer_info.update(buffer_base_pointer_info)
if any(FieldType.is_buffer(f) for f in all_fields):
resolve_buffer_accesses(ast_node, get_base_buffer_index(ast_node),
read_only_fields)
resolve_field_accesses(ast_node,
read_only_fields,
field_to_base_pointer_info=base_pointer_info)
move_constants_before_loop(ast_node)
return ast_node
def create_cuda_kernel(assignments,
function_name="kernel",
type_info=None,
indexing_creator=BlockIndexing,
iteration_slice=None,
ghost_layers=None,
skip_independence_check=False):
assert assignments, "Assignments must not be empty!"
fields_read, fields_written, assignments = add_types(assignments, type_info, not skip_independence_check)
all_fields = fields_read.union(fields_written)
read_only_fields = set([f.name for f in fields_read - fields_written])
buffers = set([f for f in all_fields if FieldType.is_buffer(f) or FieldType.is_custom(f)])
fields_without_buffers = all_fields - buffers
field_accesses = set()
num_buffer_accesses = 0
for eq in assignments:
field_accesses.update(eq.atoms(Field.Access))
field_accesses = {e for e in field_accesses if not e.is_absolute_access}
num_buffer_accesses += sum(1 for access in eq.atoms(Field.Access) if FieldType.is_buffer(access.field))
common_shape = get_common_shape(fields_without_buffers)
if iteration_slice is None:
# determine iteration slice from ghost layers
if ghost_layers is None:
# determine required number of ghost layers from field access
required_ghost_layers = max([fa.required_ghost_layers for fa in field_accesses])
ghost_layers = [(required_ghost_layers, required_ghost_layers)] * len(common_shape)
iteration_slice = []
if isinstance(ghost_layers, int):
for i in range(len(common_shape)):
iteration_slice.append(slice(ghost_layers, -ghost_layers if ghost_layers > 0 else None))
ghost_layers = [(ghost_layers, ghost_layers)] * len(common_shape)
else:
for i in range(len(common_shape)):
iteration_slice.append(slice(ghost_layers[i][0],
-ghost_layers[i][1] if ghost_layers[i][1] > 0 else None))
indexing = indexing_creator(field=list(fields_without_buffers)[0], iteration_slice=iteration_slice)
coord_mapping = indexing.coordinates
cell_idx_assignments = [SympyAssignment(LoopOverCoordinate.get_loop_counter_symbol(i), value)
for i, value in enumerate(coord_mapping)]
cell_idx_symbols = [LoopOverCoordinate.get_loop_counter_symbol(i) for i, _ in enumerate(coord_mapping)]
assignments = cell_idx_assignments + assignments
block = Block(assignments)
block = indexing.guard(block, common_shape)
unify_shape_symbols(block, common_shape=common_shape, fields=fields_without_buffers)
ast = KernelFunction(block,
Target.GPU,
Backend.CUDA,
make_python_function,
ghost_layers,
function_name,
assignments=assignments)
ast.global_variables.update(indexing.index_variables)
base_pointer_spec = [['spatialInner0']]
base_pointer_info = {f.name: parse_base_pointer_info(base_pointer_spec, [2, 1, 0],
f.spatial_dimensions, f.index_dimensions)
for f in all_fields}
coord_mapping = {f.name: cell_idx_symbols for f in all_fields}
loop_strides = list(fields_without_buffers)[0].shape
if any(FieldType.is_buffer(f) for f in all_fields):
resolve_buffer_accesses(ast, get_base_buffer_index(ast, indexing.coordinates, loop_strides), read_only_fields)
resolve_field_accesses(ast, read_only_fields, field_to_base_pointer_info=base_pointer_info,
field_to_fixed_coordinates=coord_mapping)
# add the function which determines #blocks and #threads as additional member to KernelFunction node
# this is used by the jit
# If loop counter symbols have been explicitly used in the update equations (e.g. for built in periodicity),
# they are defined here
undefined_loop_counters = {LoopOverCoordinate.is_loop_counter_symbol(s): s for s in ast.body.undefined_symbols
if LoopOverCoordinate.is_loop_counter_symbol(s) is not None}
for i, loop_counter in undefined_loop_counters.items():
ast.body.insert_front(SympyAssignment(loop_counter, indexing.coordinates[i]))
ast.indexing = indexing
return ast