misc: Prevent builtins on transient functions

devito/builtins/arithmetic.py

@@ -1,12 +1,13 @@
 import numpy as np
 
 import devito as dv
-from devito.builtins.utils import make_retval
+from devito.builtins.utils import make_retval, check_args
 
 __all__ = ['norm', 'sumall', 'sum', 'inner', 'mmin', 'mmax']
 
 
 @dv.switchconfig(log_level='ERROR')
+@check_args
 def norm(f, order=2):
     """
     Compute the norm of a Function.
@@ -41,6 +42,7 @@ def norm(f, order=2):
 
 
 @dv.switchconfig(log_level='ERROR')
+@check_args
 def sum(f, dims=None):
     """
     Compute the sum of the Function data over specified dimensions.
@@ -94,6 +96,7 @@ def sum(f, dims=None):
 
 
 @dv.switchconfig(log_level='ERROR')
+@check_args
 def sumall(f):
     """
     Compute the sum of all Function data.
@@ -123,6 +126,7 @@ def sumall(f):
 
 
 @dv.switchconfig(log_level='ERROR')
+@check_args
 def inner(f, g):
     """
     Inner product of two Functions.
@@ -177,6 +181,7 @@ def inner(f, g):
 
 
 @dv.switchconfig(log_level='ERROR')
+@check_args
 def mmin(f):
     """
     Retrieve the minimum.
@@ -200,6 +205,7 @@ def mmin(f):
 
 
 @dv.switchconfig(log_level='ERROR')
+@check_args
 def mmax(f):
     """
     Retrieve the maximum.

devito/builtins/initializers.py

@@ -2,12 +2,13 @@
 
 import devito as dv
 from devito.tools import as_tuple, as_list
-from devito.builtins.utils import nbl_to_padsize, pad_outhalo
+from devito.builtins.utils import check_args, nbl_to_padsize, pad_outhalo
 
 __all__ = ['assign', 'smooth', 'gaussian_smooth', 'initialize_function']
 
 
 @dv.switchconfig(log_level='ERROR')
+@check_args
 def assign(f, rhs=0, options=None, name='assign', assign_halo=False, **kwargs):
     """
     Assign a list of RHSs to a list of Functions.
@@ -85,6 +86,7 @@ def assign(f, rhs=0, options=None, name='assign', assign_halo=False, **kwargs):
         op(time_M=f._time_size)
 
 
+@check_args
 def smooth(f, g, axis=None):
     """
     Smooth a Function through simple moving average.
@@ -114,6 +116,7 @@ def smooth(f, g, axis=None):
         dv.Operator(dv.Eq(f, g.avg(dims=axis)), name='smoother')()
 
 
+@check_args
 def gaussian_smooth(f, sigma=1, truncate=4.0, mode='reflect'):
     """
     Gaussian smooth function.
@@ -273,6 +276,7 @@ def buff(i, j):
     return lhs, rhs, options
 
 
+@check_args
 def initialize_function(function, data, nbl, mapper=None, mode='constant',
                         name=None, pad_halo=True, **kwargs):
     """

devito/builtins/utils.py

@@ -6,7 +6,8 @@
 from devito.symbolics import uxreplace
 from devito.tools import as_tuple
 
-__all__ = ['make_retval', 'nbl_to_padsize', 'pad_outhalo', 'abstract_args']
+__all__ = ['make_retval', 'nbl_to_padsize', 'pad_outhalo', 'abstract_args',
+           'check_args']
 
 
 accumulator_mapper = {
@@ -131,3 +132,23 @@ def wrapper(*args, **kwargs):
         return func(*processed, argmap=argmap, **kwargs)
 
     return wrapper
+
+
+def check_args(func):
+    """
+    Perform checks on the arguments supplied to a builtin.
+    """
+
+    @wraps(func)
+    def wrapper(*args, **kwargs):
+        for i in args:
+            try:
+                if i.is_transient:
+                    raise ValueError(f"Cannot apply `{func.__name__}` to transient "
+                                     "function `{i.name}`")
+            except AttributeError:
+                pass
+
+        return func(*args, **kwargs)
+
+    return wrapper

tests/test_builtins.py

@@ -518,3 +518,11 @@ def test_issue_1863(self):
         assert type(v1) is np.int32
         assert type(v2) is np.float32
         assert type(v3) is np.float64
+
+    def test_is_transient(self):
+        grid = Grid(shape=(4, 4))
+
+        f = Function(name='f', grid=grid, is_transient=True)
+
+        with pytest.raises(ValueError):
+            assign(f, 4)