[Mlir-commits] [mlir] [mlir][Python] use canonical Python `isinstance` instead of `Type.isinstance` (PR #172892)

Maksim Levental llvmlistbot at llvm.org
Thu Dec 18 11:26:40 PST 2025


https://github.com/makslevental created https://github.com/llvm/llvm-project/pull/172892

We've been able to do `isinstance(x, Type)` for a quite a while now (since https://github.com/llvm/llvm-project/commit/bfb1ba752655bf09b35c486f6cc9817dbedfb1bb) so remove the special-casing in some places (and therefore support various `fp8`, `fp6`, `fp4` types). Note we could consider removing `Type.ininstance` entirely...

>From 18cc4488f39915c1c7d06718bc37f76bbcb97726 Mon Sep 17 00:00:00 2001
From: makslevental <maksim.levental at gmail.com>
Date: Thu, 18 Dec 2025 11:22:57 -0800
Subject: [PATCH] [mlir][Python] use canonical Python isinstance instead of
 Type.isinstance

---
 mlir/python/mlir/dialects/arith.py            |  30 +-----
 .../dialects/linalg/opdsl/lang/emitter.py     | 101 ++++++------------
 mlir/python/mlir/dialects/memref.py           |  15 ++-
 mlir/test/python/dialects/arith_dialect.py    |   6 +-
 4 files changed, 51 insertions(+), 101 deletions(-)

diff --git a/mlir/python/mlir/dialects/arith.py b/mlir/python/mlir/dialects/arith.py
index 88e8502a29eae..59a343435f7f0 100644
--- a/mlir/python/mlir/dialects/arith.py
+++ b/mlir/python/mlir/dialects/arith.py
@@ -21,39 +21,17 @@
     raise RuntimeError("Error loading imports from extension module") from e
 
 
-def _isa(obj: Any, cls: type):
-    try:
-        cls(obj)
-    except ValueError:
-        return False
-    return True
-
-
-def _is_any_of(obj: Any, classes: List[type]):
-    return any(_isa(obj, cls) for cls in classes)
-
-
-def _is_integer_like_type(type: Type):
-    return _is_any_of(type, [IntegerType, IndexType])
-
-
-def _is_float_type(type: Type):
-    return _is_any_of(type, [BF16Type, F16Type, F32Type, F64Type])
-
-
 @_ods_cext.register_operation(_Dialect, replace=True)
 class ConstantOp(ConstantOp):
     """Specialization for the constant op class."""
 
     @overload
-    def __init__(self, value: Attribute, *, loc=None, ip=None):
-        ...
+    def __init__(self, value: Attribute, *, loc=None, ip=None): ...
 
     @overload
     def __init__(
         self, result: Type, value: Union[int, float, _array], *, loc=None, ip=None
-    ):
-        ...
+    ): ...
 
     def __init__(self, result, value, *, loc=None, ip=None):
         if value is None:
@@ -96,9 +74,9 @@ def value(self):
 
     @property
     def literal_value(self) -> Union[int, float]:
-        if _is_integer_like_type(self.type):
+        if isinstance(self.type, (IntegerType, IndexType)):
             return IntegerAttr(self.value).value
-        elif _is_float_type(self.type):
+        elif isinstance(self.type, FloatType):
             return FloatAttr(self.value).value
         else:
             raise ValueError("only integer and float constants have literal values")
diff --git a/mlir/python/mlir/dialects/linalg/opdsl/lang/emitter.py b/mlir/python/mlir/dialects/linalg/opdsl/lang/emitter.py
index 254458a978828..a338643bad54d 100644
--- a/mlir/python/mlir/dialects/linalg/opdsl/lang/emitter.py
+++ b/mlir/python/mlir/dialects/linalg/opdsl/lang/emitter.py
@@ -412,9 +412,9 @@ def _cast(
             )
         if operand.type == to_type:
             return operand
-        if _is_integer_type(to_type):
+        if isinstance(to_type, IntegerType):
             return self._cast_to_integer(to_type, operand, is_unsigned_cast)
-        elif _is_floating_point_type(to_type):
+        elif isinstance(to_type, FloatType):
             return self._cast_to_floating_point(to_type, operand, is_unsigned_cast)
 
     def _cast_to_integer(
@@ -422,11 +422,11 @@ def _cast_to_integer(
     ) -> Value:
         to_width = IntegerType(to_type).width
         operand_type = operand.type
-        if _is_floating_point_type(operand_type):
+        if isinstance(operand_type, FloatType):
             if is_unsigned_cast:
                 return arith.FPToUIOp(to_type, operand).result
             return arith.FPToSIOp(to_type, operand).result
-        if _is_index_type(operand_type):
+        if isinstance(operand_type, IndexType):
             return arith.IndexCastOp(to_type, operand).result
         # Assume integer.
         from_width = IntegerType(operand_type).width
@@ -444,13 +444,15 @@ def _cast_to_floating_point(
         self, to_type: Type, operand: Value, is_unsigned_cast: bool
     ) -> Value:
         operand_type = operand.type
-        if _is_integer_type(operand_type):
+        if isinstance(operand_type, IntegerType):
             if is_unsigned_cast:
                 return arith.UIToFPOp(to_type, operand).result
             return arith.SIToFPOp(to_type, operand).result
         # Assume FloatType.
-        to_width = _get_floating_point_width(to_type)
-        from_width = _get_floating_point_width(operand_type)
+        assert isinstance(to_type, FloatType)
+        assert isinstance(operand_type, FloatType)
+        to_width = to_type.width
+        from_width = operand_type.width
         if to_width > from_width:
             return arith.ExtFOp(to_type, operand).result
         elif to_width < from_width:
@@ -466,89 +468,89 @@ def _type_cast_unsigned(self, type_var_name: str, operand: Value) -> Value:
         return self._cast(type_var_name, operand, True)
 
     def _unary_exp(self, x: Value) -> Value:
-        if _is_floating_point_type(x.type):
+        if isinstance(x.type, FloatType):
             return math.ExpOp(x).result
         raise NotImplementedError("Unsupported 'exp' operand: {x}")
 
     def _unary_log(self, x: Value) -> Value:
-        if _is_floating_point_type(x.type):
+        if isinstance(x.type, FloatType):
             return math.LogOp(x).result
         raise NotImplementedError("Unsupported 'log' operand: {x}")
 
     def _unary_abs(self, x: Value) -> Value:
-        if _is_floating_point_type(x.type):
+        if isinstance(x.type, FloatType):
             return math.AbsFOp(x).result
         raise NotImplementedError("Unsupported 'abs' operand: {x}")
 
     def _unary_ceil(self, x: Value) -> Value:
-        if _is_floating_point_type(x.type):
+        if isinstance(x.type, FloatType):
             return math.CeilOp(x).result
         raise NotImplementedError("Unsupported 'ceil' operand: {x}")
 
     def _unary_floor(self, x: Value) -> Value:
-        if _is_floating_point_type(x.type):
+        if isinstance(x.type, FloatType):
             return math.FloorOp(x).result
         raise NotImplementedError("Unsupported 'floor' operand: {x}")
 
     def _unary_negf(self, x: Value) -> Value:
-        if _is_floating_point_type(x.type):
+        if isinstance(x.type, FloatType):
             return arith.NegFOp(x).result
-        if _is_complex_type(x.type):
+        if isinstance(x.type, ComplexType):
             return complex.NegOp(x).result
         raise NotImplementedError("Unsupported 'negf' operand: {x}")
 
     def _binary_add(self, lhs: Value, rhs: Value) -> Value:
-        if _is_floating_point_type(lhs.type):
+        if isinstance(lhs.type, FloatType):
             return arith.AddFOp(lhs, rhs).result
-        if _is_integer_type(lhs.type) or _is_index_type(lhs.type):
+        if isinstance(lhs.type, IntegerType) or isinstance(lhs.type, IndexType):
             return arith.AddIOp(lhs, rhs).result
-        if _is_complex_type(lhs.type):
+        if isinstance(lhs.type, ComplexType):
             return complex.AddOp(lhs, rhs).result
         raise NotImplementedError("Unsupported 'add' operands: {lhs}, {rhs}")
 
     def _binary_sub(self, lhs: Value, rhs: Value) -> Value:
-        if _is_floating_point_type(lhs.type):
+        if isinstance(lhs.type, FloatType):
             return arith.SubFOp(lhs, rhs).result
-        if _is_integer_type(lhs.type) or _is_index_type(lhs.type):
+        if isinstance(lhs.type, IntegerType) or isinstance(lhs.type, IndexType):
             return arith.SubIOp(lhs, rhs).result
-        if _is_complex_type(lhs.type):
+        if isinstance(lhs.type, ComplexType):
             return complex.SubOp(lhs, rhs).result
         raise NotImplementedError("Unsupported 'sub' operands: {lhs}, {rhs}")
 
     def _binary_mul(self, lhs: Value, rhs: Value) -> Value:
-        if _is_floating_point_type(lhs.type):
+        if isinstance(lhs.type, FloatType):
             return arith.MulFOp(lhs, rhs).result
-        if _is_integer_type(lhs.type) or _is_index_type(lhs.type):
+        if isinstance(lhs.type, IntegerType) or isinstance(lhs.type, IndexType):
             return arith.MulIOp(lhs, rhs).result
-        if _is_complex_type(lhs.type):
+        if isinstance(lhs.type, ComplexType):
             return complex.MulOp(lhs, rhs).result
         raise NotImplementedError("Unsupported 'mul' operands: {lhs}, {rhs}")
 
     def _binary_max_signed(self, lhs: Value, rhs: Value) -> Value:
-        if _is_floating_point_type(lhs.type):
+        if isinstance(lhs.type, FloatType):
             return arith.MaximumFOp(lhs, rhs).result
-        if _is_integer_type(lhs.type) or _is_index_type(lhs.type):
+        if isinstance(lhs.type, IntegerType) or isinstance(lhs.type, IndexType):
             return arith.MaxSIOp(lhs, rhs).result
         raise NotImplementedError("Unsupported 'max' operands: {lhs}, {rhs}")
 
     def _binary_max_unsigned(self, lhs: Value, rhs: Value) -> Value:
-        if _is_floating_point_type(lhs.type):
+        if isinstance(lhs.type, FloatType):
             return arith.MaximumFOp(lhs, rhs).result
-        if _is_integer_type(lhs.type) or _is_index_type(lhs.type):
+        if isinstance(lhs.type, IntegerType) or isinstance(lhs.type, IndexType):
             return arith.MaxUIOp(lhs, rhs).result
         raise NotImplementedError("Unsupported 'max_unsigned' operands: {lhs}, {rhs}")
 
     def _binary_min_signed(self, lhs: Value, rhs: Value) -> Value:
-        if _is_floating_point_type(lhs.type):
+        if isinstance(lhs.type, FloatType):
             return arith.MinimumFOp(lhs, rhs).result
-        if _is_integer_type(lhs.type) or _is_index_type(lhs.type):
+        if isinstance(lhs.type, IntegerType) or isinstance(lhs.type, IndexType):
             return arith.MinSIOp(lhs, rhs).result
         raise NotImplementedError("Unsupported 'min' operands: {lhs}, {rhs}")
 
     def _binary_min_unsigned(self, lhs: Value, rhs: Value) -> Value:
-        if _is_floating_point_type(lhs.type):
+        if isinstance(lhs.type, FloatType):
             return arith.MinimumFOp(lhs, rhs).result
-        if _is_integer_type(lhs.type) or _is_index_type(lhs.type):
+        if isinstance(lhs.type, IntegerType) or isinstance(lhs.type, IndexType):
             return arith.MinUIOp(lhs, rhs).result
         raise NotImplementedError("Unsupported 'min_unsigned' operands: {lhs}, {rhs}")
 
@@ -609,40 +611,3 @@ def _add_type_mapping(
             )
     type_mapping[name] = element_or_self_type
     block_arg_types.append(element_or_self_type)
-
-
-def _is_complex_type(t: Type) -> bool:
-    return ComplexType.isinstance(t)
-
-
-def _is_floating_point_type(t: Type) -> bool:
-    # TODO: Create a FloatType in the Python API and implement the switch
-    # there.
-    return (
-        F64Type.isinstance(t)
-        or F32Type.isinstance(t)
-        or F16Type.isinstance(t)
-        or BF16Type.isinstance(t)
-    )
-
-
-def _is_integer_type(t: Type) -> bool:
-    return IntegerType.isinstance(t)
-
-
-def _is_index_type(t: Type) -> bool:
-    return IndexType.isinstance(t)
-
-
-def _get_floating_point_width(t: Type) -> int:
-    # TODO: Create a FloatType in the Python API and implement the switch
-    # there.
-    if F64Type.isinstance(t):
-        return 64
-    if F32Type.isinstance(t):
-        return 32
-    if F16Type.isinstance(t):
-        return 16
-    if BF16Type.isinstance(t):
-        return 16
-    raise NotImplementedError(f"Unhandled floating point type switch {t}")
diff --git a/mlir/python/mlir/dialects/memref.py b/mlir/python/mlir/dialects/memref.py
index c80a1b1a89358..fa60fe69a8bb8 100644
--- a/mlir/python/mlir/dialects/memref.py
+++ b/mlir/python/mlir/dialects/memref.py
@@ -7,15 +7,22 @@
 
 from ._memref_ops_gen import *
 from ._ods_common import _dispatch_mixed_values, MixedValues
-from .arith import ConstantOp, _is_integer_like_type
-from ..ir import Value, MemRefType, StridedLayoutAttr, ShapedType, Operation
+from ..ir import (
+    Value,
+    MemRefType,
+    StridedLayoutAttr,
+    ShapedType,
+    IntegerType,
+    IndexType,
+)
+from . import arith
 
 
 def _is_constant_int_like(i):
     return (
         isinstance(i, Value)
-        and isinstance(i.owner, ConstantOp)
-        and _is_integer_like_type(i.type)
+        and isinstance(i.owner, arith.ConstantOp)
+        and isinstance(i.type, (IntegerType, IndexType))
     )
 
 
diff --git a/mlir/test/python/dialects/arith_dialect.py b/mlir/test/python/dialects/arith_dialect.py
index c9af5e7b46db8..a4cfb30240231 100644
--- a/mlir/test/python/dialects/arith_dialect.py
+++ b/mlir/test/python/dialects/arith_dialect.py
@@ -42,10 +42,10 @@ def testFastMathFlags():
 def testArithValue():
     def _binary_op(lhs, rhs, op: str) -> "ArithValue":
         op = op.capitalize()
-        if arith._is_float_type(lhs.type) and arith._is_float_type(rhs.type):
+        if isinstance(lhs.type, FloatType) and isinstance(rhs.type, FloatType):
             op += "F"
-        elif arith._is_integer_like_type(lhs.type) and arith._is_integer_like_type(
-            lhs.type
+        elif isinstance(lhs.type, (IntegerType, IndexType)) and isinstance(
+            lhs.type, (IntegerType, IndexType)
         ):
             op += "I"
         else:



More information about the Mlir-commits mailing list