Misc fixes for MacOS. (llvm#255)

* Change aligned_alloc -> malloc. It can fail (and does on MacOS) and is a bit over-aggressive optimization for a reference backend.
* Fixed a fragile test that prints -0.0 on MacOS.
* Fail the test (not the framework) on failure to trace (Torch on MacOS is missing features).
* Fix .so -> .dylib for compiler runtime.

frontends/pytorch/python/torch_mlir/torchscript/e2e_test/framework.py

@@ -227,9 +227,9 @@ def run_tests(tests: List[Test], config: TestConfig) -> List[TestResult]:
     """Invoke the given `Test`'s with the provided `TestConfig`."""
     results = []
     for test in tests:
-        golden_trace = _generate_golden_trace(test)
         # TODO: Precompile everything in parallel.
         try:
+            golden_trace = _generate_golden_trace(test)
             compiled = config.compile(test.program_factory())
         except Exception as e:
             results.append(

lib/RefBackend/Runtime/Runtime.cpp

@@ -488,7 +488,8 @@ RtValue refbackrt::createRtValueFromOutputArgInfo(const OutputArgInfo &info) {
     switch (info.elementType) {
     case ElementType::F32: {
       auto byteSize = numel * sizeof(float);
-      data = static_cast<void *>(aligned_alloc(32, byteSize));
+      data = static_cast<void *>(malloc(byteSize));
+      assert(data && "could not allocate tensor");
       memset(data, 0, byteSize);
       return RtValue(Tensor::create(shape, ElementType::F32, data));
       break;

python/npcomp/compiler/generic/backend/refjit.py

@@ -3,6 +3,7 @@
 #  SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 
 import os
+import platform
 
 _refjit = None
 
@@ -40,7 +41,11 @@ def is_enabled() -> bool:
 def get_runtime_libs():
   # The _refjit_resources directory is at the npcomp.compiler level.
   resources_dir = os.path.join(os.path.dirname(__file__))
-  return [os.path.join(resources_dir, "libNPCOMPCompilerRuntimeShlib.so")]
+  suffix = ".so"
+  if platform.system() == "Darwin":
+    suffix = ".dylib"
+  shlib_name = f"libNPCOMPCompilerRuntimeShlib{suffix}"
+  return [os.path.join(resources_dir, shlib_name)]
 
 
 class JitModuleInvoker:

python/npcomp/types.py

@@ -65,7 +65,7 @@ class _LiterateEnum(Enum):
     Traceback (most recent call last):
     ...
     ValueError: Cannot parse SampleEnum 1.0
-    
+
   """
 
   @classmethod
@@ -111,11 +111,11 @@ class TypeClass(_LiterateEnum):
 
 class ValueType:
   """The type a value can take in the npcomp language.
-  
+
   Types of values in npcomp are always being refined and are therefore
-  mutable. Instances represent the type derived for a single value, not a 
+  mutable. Instances represent the type derived for a single value, not a
   concept of "typeness" generally.
-  
+
     >>> ValueType()
     Any
     >>> ValueType('NdArray')
@@ -166,7 +166,7 @@ def constraints(self):
 
 class ValueTypeList:
   """Models a list of ValueTypes.
-  
+
     >>> v3 = ValueTypeList(3)
     >>> v3
     (Any, Any, Any)
@@ -178,7 +178,7 @@ class ValueTypeList:
     >>> v3[2] += Rank(2)
     >>> v3
     (Any, Any, NdArray[Rank(2)])
-    
+
   With names:
     >>> v3 = ValueTypeList(3, [None, "b", None])
     >>> v3[1] = 'NdArray'
@@ -221,11 +221,11 @@ def __repr__(self):
 
 class Signature:
   """A function signature.
-    
+
     This currently only models a linear list of positional arguments and
     assumes that multiple results will be represented by some form of tuple
     type.
-      
+
       >>> Signature()
       () -> Any
       >>> Signature(2)
@@ -279,7 +279,7 @@ def __repr__(self):
 
 class ArrayParams:
   """Represents parameters defining how to construct an array.
-  
+
     >>> ArrayParams()
     ArrayParams(dtype=Unspec)
     >>> ArrayParams(np.float32)
@@ -309,34 +309,34 @@ def rank(self):
   @classmethod
   def from_constraints(cls, constraints):
     """Constructs params for a TypeConstraints list.
-    
+
     Unconstrained:
       >>> ArrayParams.from_constraints(TypeConstraints())
       ArrayParams(dtype=Unspec)
-      
+
     DType constrained:
       >>> ArrayParams.from_constraints(TypeConstraints(DType(np.float32)))
       ArrayParams(dtype=float32)
 
-    Rank constrained:      
+    Rank constrained:
       >>> ArrayParams.from_constraints(TypeConstraints(Rank(2)))
       ArrayParams(dtype=Unspec, shape=(-1, -1))
-      
+
     Shape constrained:
       >>> ArrayParams.from_constraints(TypeConstraints(Shape(1, 2, 3)))
       ArrayParams(dtype=Unspec, shape=(1, 2, 3))
       >>> ArrayParams.from_constraints(TypeConstraints(
       ...   Rank(3), Shape(1, 2, 3)))
       ArrayParams(dtype=Unspec, shape=(1, 2, 3))
-      
+
     Shape constrained with dynamic dim constraint:
       >>> ArrayParams.from_constraints(TypeConstraints(
       ...   Shape(1, 2, 3), DynamicDim(1)))
       ArrayParams(dtype=Unspec, shape=(1, -1, 3))
       >>> ArrayParams.from_constraints(TypeConstraints(
       ...   Shape(1, 2, 3), DynamicDim((0, 2))))
       ArrayParams(dtype=Unspec, shape=(-1, 2, -1))
-      
+
     Errors:
       >>> ArrayParams.from_constraints(TypeConstraints(
       ...   Rank(4), Shape(1, 2, 3)))
@@ -346,7 +346,7 @@ def from_constraints(cls, constraints):
       >>> ArrayParams.from_constraints(TypeConstraints(
       ...   Shape(1, 2, 3), DynamicDim((0, 5))))
       Traceback (most recent call last):
-      ...      
+      ...
       ValueError: Out of range DimFlag(Dynamic, (0, 5)) for shape [-1, 2, 3]
     """
     # TODO: Should have a 'canonicalize' method on TypeConstraints which
@@ -395,7 +395,7 @@ def __repr__(self):
   @property
   def is_concrete(self):
     """Returns true if the parameters are sufficient to construct an ndarray.
-    
+
       >>> ArrayParams().is_concrete
       False
       >>> ArrayParams(dtype=np.float32).is_concrete
@@ -417,26 +417,26 @@ def is_concrete(self):
   def mlir_tensor_type_asm(self):
     """Get a corresponding MLIR tensor type.
 
-    Fully Unspecified:    
+    Fully Unspecified:
       >>> ArrayParams().mlir_tensor_type_asm
       'tensor<*x!numpy.any_dtype>'
-      
+
     Unranked:
       >>> ArrayParams(dtype=np.float32).mlir_tensor_type_asm
       'tensor<*xf32>'
-      
+
     Ranked:
       >>> ArrayParams(dtype=np.float32, rank=3).mlir_tensor_type_asm
       'tensor<?x?x?xf32>'
       >>> ArrayParams(dtype=np.float32, shape=(-1, -1)).mlir_tensor_type_asm
       'tensor<?x?xf32>'
-      
+
     Scalar:
       >>> ArrayParams(dtype=np.float32, rank=0).mlir_tensor_type_asm
       'tensor<f32>'
       >>> ArrayParams(dtype=np.float32, shape=()).mlir_tensor_type_asm
       'tensor<f32>'
-      
+
     Shaped:
       >>> ArrayParams(dtype=np.float32, shape=(2, 3)).mlir_tensor_type_asm
       'tensor<2x3xf32>'
@@ -460,12 +460,12 @@ def mlir_tensor_type_asm(self):
 
   def new_ndarray(self):
     """Creates a new ndarray from these params.
-    
+
       >>> ArrayParams().new_ndarray()
       Traceback (most recent call last):
       ...
       ValueError: ArrayParams(dtype=Unspec) is not concrete
-      >>> ArrayParams(np.float32, (1, 2)).new_ndarray() * 0.0
+      >>> (ArrayParams(np.float32, (1, 2)).new_ndarray() * 0.0 + 1.0) * 0.0
       array([[0., 0.]], dtype=float32)
     """
     if not self.is_concrete:
@@ -480,7 +480,7 @@ class TypeConstraint:
 
 class TypeConstraints(list):
   """Collection of type constraints.
-  
+
     >>> TypeConstraints([DynamicDim()])
     TypeConstraints(DimFlag(Dynamic, Unspec))
     >>> TypeConstraints([DynamicDim(), Rank(4)])
@@ -554,9 +554,9 @@ def dim_flag(self):
 
 class DType(ArrayConstraint):
   """A constraint on a dtype.
-  
+
   DType constraints are exclusive with only one permitted in a set.
-  
+
     >>> DType(np.float32)
     DType(float32)
     >>> DType("foobar")
@@ -597,7 +597,7 @@ class Rank(ArrayConstraint):
     Traceback (most recent call last):
     ...
     AssertionError
-    
+
   """
   __slots__ = ["_rank"]
 
@@ -619,9 +619,9 @@ def __repr__(self):
 
 class Shape(ArrayConstraint):
   """Establishes a static shape for an array.
-  
+
   All dimensions must be a non-negative integer or Unspec.
-  
+
     >>> Shape(1, 2, 3)
     Shape(1, 2, 3)
     >>> Shape(Unspec, 1)
@@ -665,9 +665,9 @@ class DimFlagEnum(_LiterateEnum):
 
 class DimFlag(ArrayConstraint):
   """Generic flag applying to one or more dimensions.
-  
+
   If dims is Unspec, the flag applies to all dims.
-  
+
     >>> DimFlag("Dynamic")
     DimFlag(Dynamic, Unspec)
     >>> DimFlag("Dynamic", 1)