psyclone.transformations.DynamoOMPParallelLoopTrans Class Reference

Inheritance diagram for psyclone.transformations.DynamoOMPParallelLoopTrans:

Collaboration diagram for psyclone.transformations.DynamoOMPParallelLoopTrans:

Public Member Functions
def	__init__ (self, omp_directive="do", omp_schedule="static")
def	__str__ (self)
def	validate (self, node, options=None)
Public Member Functions inherited from psyclone.transformations.OMPParallelLoopTrans
def	apply (self, node, options=None)
Public Member Functions inherited from psyclone.psyir.transformations.omp_loop_trans.OMPLoopTrans
def	omp_directive (self)
def	omp_directive (self, value)
def	omp_schedule (self)
def	omp_schedule (self, value)
Public Member Functions inherited from psyclone.psyir.transformations.loop_trans.LoopTrans
def	name (self)

Additional Inherited Members
Public Attributes inherited from psyclone.psyir.transformations.omp_loop_trans.OMPLoopTrans
	omp_schedule
	omp_directive
Static Public Attributes inherited from psyclone.psyir.transformations.parallel_loop_trans.ParallelLoopTrans
tuple	excluded_node_types = (nodes.Return, psyGen.HaloExchange, nodes.CodeBlock)
Static Public Attributes inherited from psyclone.psyir.transformations.loop_trans.LoopTrans
tuple	excluded_node_types = ()

Detailed Description

 Dynamo-specific OpenMP loop transformation. Adds Dynamo specific
    validity checks. Actual transformation is done by the
    :py:class:`base class <OMPParallelLoopTrans>`.

    :param str omp_directive: choose which OpenMP loop directive to use.
        Defaults to "do".
    :param str omp_schedule: the OpenMP schedule to use. Must be one of
        'runtime', 'static', 'dynamic', 'guided' or 'auto'. Defaults to
        'static'.

Definition at line 740 of file transformations.py.

Member Function Documentation

validate()

def psyclone.transformations.DynamoOMPParallelLoopTrans.validate	(		self,
			node,
			options = None
	)

Perform LFRic-specific loop validity checks then call the `validate`
method of the base class.

:param node: the Node in the Schedule to check
:type node: :py:class:`psyclone.psyir.nodes.Node`
:param options: a dictionary with options for transformations.
:type options: Optional[Dict[str, Any]]

:raises TransformationError: if the supplied Node is not a LFRicLoop.
:raises TransformationError: if the associated loop requires
    colouring.

Reimplemented from psyclone.psyir.transformations.parallel_loop_trans.ParallelLoopTrans.

Definition at line 760 of file transformations.py.

    def validate(self, node, options=None):
        '''
        Perform LFRic-specific loop validity checks then call the `validate`
        method of the base class.
 
        :param node: the Node in the Schedule to check
        :type node: :py:class:`psyclone.psyir.nodes.Node`
        :param options: a dictionary with options for transformations.
        :type options: Optional[Dict[str, Any]]
 
        :raises TransformationError: if the supplied Node is not a LFRicLoop.
        :raises TransformationError: if the associated loop requires
            colouring.
        '''
        if not isinstance(node, LFRicLoop):
            raise TransformationError(
                f"Error in {self.name} transformation. The supplied node "
                f"must be a LFRicLoop but got '{type(node).__name__}'")
 
        # If the loop is not already coloured then check whether or not
        # it should be. If the field space is discontinuous (including
        # any_discontinuous_space) then we don't need to worry about
        # colouring.
        const = LFRicConstants()
        if node.field_space.orig_name not in const.VALID_DISCONTINUOUS_NAMES:
            if node.loop_type != 'colour' and node.has_inc_arg():
                raise TransformationError(
                    f"Error in {self.name} transformation. The kernel has an "
                    f"argument with INC access. Colouring is required.")
        # As this is a domain-specific loop, we don't perform general
        # dependence analysis because it is too conservative and doesn't
        # account for the special steps taken for such a loop at code-
        # generation time (e.g. the way we ensure variables are given the
        # correct sharing attributes).
        local_options = options.copy() if options else {}
        local_options["force"] = True
        super().validate(node, options=local_options)
 
 

Here is the caller graph for this function:

---

The documentation for this class was generated from the following file:

• /home/docs/checkouts/readthedocs.org/user_builds/psyclone-ref/checkouts/latest/src/psyclone/transformations.py