def statement_evaluate(leaf, ctx):
expr = leaf.expression
if isinstance(expr, gem.ListTensor):
ops = []
var, index = ctx.pymbolic_variable_and_destruct(expr)
for multiindex, value in numpy.ndenumerate(expr.array):
ops.append(
lp.Assignment(p.Subscript(var, index + multiindex),
expression(value, ctx),
within_inames=ctx.active_inames()))
return ops
elif isinstance(expr, gem.Constant):
return []
elif isinstance(expr, gem.ComponentTensor):
idx = ctx.gem_to_pym_multiindex(expr.multiindex)
var, sub_idx = ctx.pymbolic_variable_and_destruct(expr)
lhs = p.Subscript(var, idx + sub_idx)
with active_indices(dict(zip(expr.multiindex, idx)),
ctx) as ctx_active:
return [
lp.Assignment(lhs,
expression(expr.children[0], ctx_active),
within_inames=ctx_active.active_inames())
]
elif isinstance(expr, gem.Inverse):
idx = ctx.pymbolic_multiindex(expr.shape)
var = ctx.pymbolic_variable(expr)
lhs = (SubArrayRef(idx, p.Subscript(var, idx)), )
idx_reads = ctx.pymbolic_multiindex(expr.children[0].shape)
var_reads = ctx.pymbolic_variable(expr.children[0])
reads = (SubArrayRef(idx_reads, p.Subscript(var_reads, idx_reads)), )
rhs = p.Call(p.Variable("inverse"), reads)
return [
lp.CallInstruction(lhs, rhs, within_inames=ctx.active_inames())
]
elif isinstance(expr, gem.Solve):
idx = ctx.pymbolic_multiindex(expr.shape)
var = ctx.pymbolic_variable(expr)
lhs = (SubArrayRef(idx, p.Subscript(var, idx)), )
reads = []
for child in expr.children:
idx_reads = ctx.pymbolic_multiindex(child.shape)
var_reads = ctx.pymbolic_variable(child)
reads.append(
SubArrayRef(idx_reads, p.Subscript(var_reads, idx_reads)))
rhs = p.Call(p.Variable("solve"), tuple(reads))
return [
lp.CallInstruction(lhs, rhs, within_inames=ctx.active_inames())
]
else:
return [
lp.Assignment(ctx.pymbolic_variable(expr),
expression(expr, ctx, top=True),
within_inames=ctx.active_inames())
]
def slate_call(self, kernel):
# Slate kernel call
call = pym.Call(pym.Variable(kernel.name), tuple())
output_var = pym.Variable(kernel.args[0].name)
slate_kernel_call_output = self.generate_lhs(self.expression, output_var)
insn = loopy.CallInstruction((slate_kernel_call_output,), call, id="slate_kernel_call")
return [insn]
def generate_tsfc_calls(self, terminal, loopy_tensor):
"""A setup method to initialize all the local assembly
kernels generated by TSFC. This function also collects any
information regarding orientations and extra include directories.
"""
cxt_kernels = self.tsfc_cxt_kernels(terminal)
for cxt_kernel in cxt_kernels:
for tsfc_kernel in cxt_kernel.tsfc_kernels:
integral_type = cxt_kernel.original_integral_type
slate_tensor = cxt_kernel.tensor
mesh = slate_tensor.ufl_domain()
kinfo = tsfc_kernel.kinfo
reads = []
inames_dep = []
if integral_type not in self.supported_integral_types:
raise ValueError("Integral type '%s' not recognized" %
integral_type)
# Prepare lhs and args for call to tsfc kernel
output_var = pym.Variable(loopy_tensor.name)
reads.append(output_var)
output = self.generate_lhs(slate_tensor, output_var)
kernel_data = self.collect_tsfc_kernel_data(
mesh, cxt_kernel.coefficients, self.bag.coefficients,
kinfo)
reads.extend(self.loopify_tsfc_kernel_data(kernel_data))
# Generate predicates for different integral types
if self.is_integral_type(integral_type, "cell_integral"):
predicates = None
if kinfo.subdomain_id != "otherwise":
raise NotImplementedError(
"No subdomain markers for cells yet")
elif self.is_integral_type(integral_type, "facet_integral"):
predicates, fidx, facet_arg = self.facet_integral_predicates(
mesh, integral_type, kinfo)
reads.append(facet_arg)
inames_dep.append(fidx[0].name)
elif self.is_integral_type(integral_type, "layer_integral"):
predicates = self.layer_integral_predicates(
slate_tensor, integral_type)
else:
raise ValueError(
"Unhandled integral type {}".format(integral_type))
# TSFC kernel call
key = self.bag.call_name_generator(integral_type)
call = pym.Call(pym.Variable(kinfo.kernel.name), tuple(reads))
insn = loopy.CallInstruction(
(output, ),
call,
within_inames=frozenset(inames_dep),
predicates=predicates,
id=key)
yield insn, kinfo.kernel.code
def slate_call(self, kernel, temporaries):
# Slate kernel call
reads = []
for t in temporaries:
shape = t.shape
name = t.name
idx = self.bag.index_creator(shape)
reads.append(SubArrayRef(idx, pym.Subscript(pym.Variable(name), idx)))
call = pym.Call(pym.Variable(kernel.name), tuple(reads))
output_var = pym.Variable(kernel.args[0].name)
slate_kernel_call_output = self.generate_lhs(self.expression, output_var)
insn = loopy.CallInstruction((slate_kernel_call_output,), call, id="slate_kernel_call")
return insn
def test_call_with_no_returned_value(ctx_factory):
import pymbolic.primitives as p
ctx = ctx_factory()
queue = cl.CommandQueue(ctx)
knl = lp.make_kernel("{:}",
[lp.CallInstruction((), p.Call(p.Variable("f"), ()))])
from library_for_test import no_ret_f_mangler, no_ret_f_preamble_gen
knl = lp.register_function_manglers(knl, [no_ret_f_mangler])
knl = lp.register_preamble_generators(knl, [no_ret_f_preamble_gen])
evt, _ = knl(queue)
def statement_functioncall(expr, context):
parameters = context.parameters
free_indices = set(i.name for i in expr.free_indices)
writes = []
reads = []
for access, child in zip(expr.access, expr.children):
var = expression(child, parameters)
if isinstance(var, pym.Subscript):
# tensor argument
indices = []
sweeping_indices = []
for index in var.index_tuple:
indices.append(index)
if isinstance(index,
pym.Variable) and index.name in free_indices:
sweeping_indices.append(index)
arg = SubArrayRef(tuple(sweeping_indices), var)
else:
# scalar argument or constant
arg = var
if access is READ or (isinstance(child, Argument)
and isinstance(child.dtype, OpaqueType)):
reads.append(arg)
else:
writes.append(arg)
within_inames = context.within_inames[expr]
predicates = frozenset(context.conditions)
id, depends_on = context.instruction_dependencies[expr]
call = pym.Call(pym.Variable(expr.name), tuple(reads))
return loopy.CallInstruction(tuple(writes),
call,
within_inames=within_inames,
predicates=predicates,
id=id,
depends_on=depends_on,
depends_on_is_final=True)