Skip to content
Snippets Groups Projects
vectorization.py 10.5 KiB
Newer Older
import sympy as sp
import warnings
Martin Bauer's avatar
Martin Bauer committed
from typing import Union, Container
from pystencils.backends.simd_instruction_sets import get_vector_instruction_set
from pystencils.fast_approximation import fast_division, fast_sqrt, fast_inv_sqrt
from pystencils.integer_functions import modulo_floor, modulo_ceil
Martin Bauer's avatar
Martin Bauer committed
from pystencils.sympyextensions import fast_subs
Martin Bauer's avatar
Martin Bauer committed
from pystencils.data_types import TypedSymbol, VectorType, get_type_of_expression, vector_memory_access, cast_func, \
    collate_types, PointerType
import pystencils.astnodes as ast
from pystencils.transformations import cut_loop, filtered_tree_iteration, replace_inner_stride_with_one
Martin Bauer's avatar
Martin Bauer committed
from pystencils.field import Field


def vectorize(kernel_ast: ast.KernelFunction, instruction_set: str = 'avx',
              assume_aligned: bool = False, nontemporal: Union[bool, Container[Union[str, Field]]] = False,
              assume_inner_stride_one: bool = False, assume_sufficient_line_padding: bool = True):
Martin Bauer's avatar
Martin Bauer committed
    """Explicit vectorization using SIMD vectorization via intrinsics.

    Args:
        kernel_ast: abstract syntax tree (KernelFunction node)
        instruction_set: one of the supported vector instruction sets, currently ('sse', 'avx' and 'avx512')
Martin Bauer's avatar
Martin Bauer committed
        assume_aligned: assume that the first inner cell of each line is aligned. If false, only unaligned-loads are
                        used. If true, some of the loads are assumed to be from aligned memory addresses.
                        For example if x is the fastest coordinate, the access to center can be fetched via an
                        aligned-load instruction, for the west or east accesses potentially slower unaligend-load
                        instructions have to be used.
        nontemporal: a container of fields or field names for which nontemporal (streaming) stores are used.
                     If true, nontemporal access instructions are used for all fields.
        assume_inner_stride_one: kernels with non-constant inner loop bound and strides can not be vectorized since
                                 the inner loop stride is a runtime variable and thus might not be always 1.
                                 If this parameter is set to true, the inner stride is assumed to be always one.
                                 This has to be ensured at runtime!
        assume_sufficient_line_padding: if True and assume_inner_stride_one, no tail loop is created but loop is
                                        extended by at most (vector_width-1) elements
                                        assumes that at the end of each line there is enough padding with dummy data
                                        depending on the access pattern there might be additional padding
                                        required at the end of the array
Martin Bauer's avatar
Martin Bauer committed
    """
    if instruction_set is None:
        return
    
Martin Bauer's avatar
Martin Bauer committed
    all_fields = kernel_ast.fields_accessed
    if nontemporal is None or nontemporal is False:
        nontemporal = {}
    elif nontemporal is True:
        nontemporal = all_fields

    if assume_inner_stride_one:
        replace_inner_stride_with_one(kernel_ast)

    field_float_dtypes = set(f.dtype for f in all_fields if f.dtype.is_float())
Martin Bauer's avatar
Martin Bauer committed
    if len(field_float_dtypes) != 1:
        raise NotImplementedError("Cannot vectorize kernels that contain accesses "
                                  "to differently typed floating point fields")
    float_size = field_float_dtypes.pop().numpy_dtype.itemsize
    assert float_size in (8, 4)
    vector_is = get_vector_instruction_set('double' if float_size == 8 else 'float',
                                           instruction_set=instruction_set)
Martin Bauer's avatar
Martin Bauer committed
    vector_width = vector_is['width']
    kernel_ast.instruction_set = vector_is

    vectorize_inner_loops_and_adapt_load_stores(kernel_ast, vector_width, assume_aligned,
                                                nontemporal, assume_sufficient_line_padding)
Martin Bauer's avatar
Martin Bauer committed
    insert_vector_casts(kernel_ast)


def vectorize_inner_loops_and_adapt_load_stores(ast_node, vector_width, assume_aligned, nontemporal_fields,
                                                assume_sufficient_line_padding):
Martin Bauer's avatar
Martin Bauer committed
    """Goes over all innermost loops, changes increment to vector width and replaces field accesses by vector type."""
Martin Bauer's avatar
Martin Bauer committed
    all_loops = filtered_tree_iteration(ast_node, ast.LoopOverCoordinate, stop_type=ast.SympyAssignment)
    inner_loops = [n for n in all_loops if n.is_innermost_loop]
    zero_loop_counters = {l.loop_counter_symbol: 0 for l in all_loops}
Martin Bauer's avatar
Martin Bauer committed
    for loop_node in inner_loops:
        loop_range = loop_node.stop - loop_node.start
Martin Bauer's avatar
Martin Bauer committed
        # cut off loop tail, that is not a multiple of four
        if assume_aligned and assume_sufficient_line_padding:
            loop_range = loop_node.stop - loop_node.start
            new_stop = loop_node.start + modulo_ceil(loop_range, vector_width)
            loop_node.stop = new_stop
        else:
            cutting_point = modulo_floor(loop_range, vector_width) + loop_node.start
            loop_nodes = cut_loop(loop_node, [cutting_point])
            assert len(loop_nodes) in (1, 2)  # 2 for main and tail loop, 1 if loop range divisible by vector width
            loop_node = loop_nodes[0]
        # Find all array accesses (indexed) that depend on the loop counter as offset
Martin Bauer's avatar
Martin Bauer committed
        loop_counter_symbol = ast.LoopOverCoordinate.get_loop_counter_symbol(loop_node.coordinate_to_loop_over)
        substitutions = {}
        successful = True
Martin Bauer's avatar
Martin Bauer committed
        for indexed in loop_node.atoms(sp.Indexed):
Martin Bauer's avatar
Martin Bauer committed
            if loop_counter_symbol in index.atoms(sp.Symbol):
                loop_counter_is_offset = loop_counter_symbol not in (index - loop_counter_symbol).atoms()
Martin Bauer's avatar
Martin Bauer committed
                aligned_access = (index - loop_counter_symbol).subs(zero_loop_counters) == 0
Martin Bauer's avatar
Martin Bauer committed
                if not loop_counter_is_offset:
Martin Bauer's avatar
Martin Bauer committed
                typed_symbol = base.label
                assert type(typed_symbol.dtype) is PointerType, \
                    "Type of access is {}, {}".format(typed_symbol.dtype, indexed)
Martin Bauer's avatar
Martin Bauer committed

                vec_type = VectorType(typed_symbol.dtype.base_type, vector_width)
                use_aligned_access = aligned_access and assume_aligned
                nontemporal = False
                if hasattr(indexed, 'field'):
                    nontemporal = (indexed.field in nontemporal_fields) or (indexed.field.name in nontemporal_fields)
                substitutions[indexed] = vector_memory_access(indexed, vec_type, use_aligned_access, nontemporal)
        if not successful:
            warnings.warn("Could not vectorize loop because of non-consecutive memory access")
            continue
Martin Bauer's avatar
Martin Bauer committed
        loop_node.step = vector_width
        loop_node.subs(substitutions)
Martin Bauer's avatar
Martin Bauer committed
def insert_vector_casts(ast_node):
    """Inserts necessary casts from scalar values to vector values."""

    handled_functions = (sp.Add, sp.Mul, fast_division, fast_sqrt, fast_inv_sqrt)

Martin Bauer's avatar
Martin Bauer committed
    def visit_expr(expr):
        if isinstance(expr, cast_func) or isinstance(expr, vector_memory_access):
Martin Bauer's avatar
Martin Bauer committed
            return expr
        elif expr.func in handled_functions or isinstance(expr, sp.Rel) or isinstance(expr, sp.boolalg.BooleanFunction):
Martin Bauer's avatar
Martin Bauer committed
            new_args = [visit_expr(a) for a in expr.args]
            arg_types = [get_type_of_expression(a) for a in new_args]
            if not any(type(t) is VectorType for t in arg_types):
                return expr
            else:
Martin Bauer's avatar
Martin Bauer committed
                target_type = collate_types(arg_types)
Martin Bauer's avatar
Martin Bauer committed
                casted_args = [cast_func(a, target_type) if t != target_type else a
Martin Bauer's avatar
Martin Bauer committed
                               for a, t in zip(new_args, arg_types)]
                return expr.func(*casted_args)
        elif expr.func is sp.Pow:
Martin Bauer's avatar
Martin Bauer committed
            new_arg = visit_expr(expr.args[0])
            return expr.func(new_arg, expr.args[1])
        elif expr.func == sp.Piecewise:
Martin Bauer's avatar
Martin Bauer committed
            new_results = [visit_expr(a[0]) for a in expr.args]
            new_conditions = [visit_expr(a[1]) for a in expr.args]
            types_of_results = [get_type_of_expression(a) for a in new_results]
            types_of_conditions = [get_type_of_expression(a) for a in new_conditions]
Martin Bauer's avatar
Martin Bauer committed
            result_target_type = get_type_of_expression(expr)
            condition_target_type = collate_types(types_of_conditions)
            if type(condition_target_type) is VectorType and type(result_target_type) is not VectorType:
                result_target_type = VectorType(result_target_type, width=condition_target_type.width)
            if type(condition_target_type) is not VectorType and type(result_target_type) is VectorType:
                condition_target_type = VectorType(condition_target_type, width=result_target_type.width)
Martin Bauer's avatar
Martin Bauer committed
            casted_results = [cast_func(a, result_target_type) if t != result_target_type else a
Martin Bauer's avatar
Martin Bauer committed
                              for a, t in zip(new_results, types_of_results)]
Martin Bauer's avatar
Martin Bauer committed
            casted_conditions = [cast_func(a, condition_target_type)
Martin Bauer's avatar
Martin Bauer committed
                                 if t != condition_target_type and a is not True else a
                                 for a, t in zip(new_conditions, types_of_conditions)]
Martin Bauer's avatar
Martin Bauer committed
            return sp.Piecewise(*[(r, c) for r, c in zip(casted_results, casted_conditions)])
Martin Bauer's avatar
Martin Bauer committed
    def visit_node(node, substitution_dict):
        substitution_dict = substitution_dict.copy()
        for arg in node.args:
            if isinstance(arg, ast.SympyAssignment):
                assignment = arg
                subs_expr = fast_subs(assignment.rhs, substitution_dict,
                                      skip=lambda e: isinstance(e, ast.ResolvedFieldAccess))
                assignment.rhs = visit_expr(subs_expr)
                rhs_type = get_type_of_expression(assignment.rhs)
                if isinstance(assignment.lhs, TypedSymbol):
                    lhs_type = assignment.lhs.dtype
                    if type(rhs_type) is VectorType and type(lhs_type) is not VectorType:
                        new_lhs_type = VectorType(lhs_type, rhs_type.width)
                        new_lhs = TypedSymbol(assignment.lhs.name, new_lhs_type)
                        substitution_dict[assignment.lhs] = new_lhs
                        assignment.lhs = new_lhs
Martin Bauer's avatar
Martin Bauer committed
                elif isinstance(assignment.lhs.func, cast_func):
Martin Bauer's avatar
Martin Bauer committed
                    lhs_type = assignment.lhs.args[1]
                    if type(lhs_type) is VectorType and type(rhs_type) is not VectorType:
                        assignment.rhs = cast_func(assignment.rhs, lhs_type)
            else:
                visit_node(arg, substitution_dict)

    visit_node(ast_node, {})