def build_compile_and_run_SpMM(attr: st.EncodingAttr, support_lib: str, compiler): # Build. module = build_SpMM(attr) func = str(module.operation.regions[0].blocks[0].operations[0].operation) module = ir.Module.parse(func + boilerplate(attr)) # Compile. compiler(module) engine = execution_engine.ExecutionEngine(module, opt_level=0, shared_libs=[support_lib]) # Set up numpy input, invoke the kernel, and get numpy output. # Built-in bufferization uses in-out buffers. # TODO: replace with inplace comprehensive bufferization. Cin = np.zeros((3, 2), np.double) Cout = np.zeros((3, 2), np.double) Cin_memref_ptr = ctypes.pointer( ctypes.pointer(rt.get_ranked_memref_descriptor(Cin))) Cout_memref_ptr = ctypes.pointer( ctypes.pointer(rt.get_ranked_memref_descriptor(Cout))) engine.invoke('main', Cout_memref_ptr, Cin_memref_ptr) Cresult = rt.ranked_memref_to_numpy(Cout_memref_ptr[0]) # Sanity check on computed result. expected = [[12.3, 12.0], [0.0, 0.0], [16.5, 19.8]] if np.allclose(Cresult, expected): pass else: quit(f'FAILURE')
def build_compile_and_run_SDDMMM(attr: st.EncodingAttr, opt: str, support_lib: str, compiler): # Build. module = build_SDDMM(attr) func = str(module.operation.regions[0].blocks[0].operations[0].operation) module = ir.Module.parse(func + boilerplate(attr)) # Compile. compiler(module) engine = execution_engine.ExecutionEngine(module, opt_level=0, shared_libs=[support_lib]) # Set up numpy input and buffer for output. a = np.array([[1.1, 2.1, 3.1, 4.1, 5.1, 6.1, 7.1, 8.1], [1.2, 2.2, 3.2, 4.2, 5.2, 6.2, 7.2, 8.2], [1.3, 2.3, 3.3, 4.3, 5.3, 6.3, 7.3, 8.3], [1.4, 2.4, 3.4, 4.4, 5.4, 6.4, 7.4, 8.4], [1.5, 2.5, 3.5, 4.5, 5.5, 6.5, 7.5, 8.5], [1.6, 2.6, 3.6, 4.6, 5.6, 6.6, 7.6, 8.6], [1.7, 2.7, 3.7, 4.7, 5.7, 6.7, 7.7, 8.7], [1.8, 2.8, 3.8, 4.8, 5.8, 6.8, 7.8, 8.8]], np.float64) b = np.ones((8, 8), np.float64) c = np.zeros((8, 8), np.float64) mem_a = ctypes.pointer(ctypes.pointer(rt.get_ranked_memref_descriptor(a))) mem_b = ctypes.pointer(ctypes.pointer(rt.get_ranked_memref_descriptor(b))) mem_c = ctypes.pointer(ctypes.pointer(rt.get_ranked_memref_descriptor(c))) # Allocate a MemRefDescriptor to receive the output tensor. # The buffer itself is allocated inside the MLIR code generation. ref_out = rt.make_nd_memref_descriptor(2, ctypes.c_double)() mem_out = ctypes.pointer(ctypes.pointer(ref_out)) # Invoke the kernel and get numpy output. # Built-in bufferization uses in-out buffers. # TODO: replace with inplace comprehensive bufferization. engine.invoke('main', mem_out, mem_a, mem_b, mem_c) # Sanity check on computed result. Only a few elements # are sampled from the full dense matrix multiplication. full_matmul = np.matmul(a, b) expected = np.zeros((8, 8), np.float64) expected[0, 0] = 1.0 * full_matmul[0, 0] expected[0, 2] = 2.0 * full_matmul[0, 2] expected[4, 1] = 3.0 * full_matmul[4, 1] c = rt.ranked_memref_to_numpy(mem_out[0]) if np.allclose(c, expected): pass else: quit(f'FAILURE')
def run(self, np_arg0: np.ndarray) -> np.ndarray: """Runs the test on the given numpy array, and returns the resulting numpy array.""" assert self._engine is not None, \ 'StressTest: must call compile() before run()' self._assertEqualsRoundtripTp( self._tyconv.get_RankedTensorType_of_nparray(np_arg0)) np_out = np.zeros(np_arg0.shape, dtype=np_arg0.dtype) self._assertEqualsRoundtripTp( self._tyconv.get_RankedTensorType_of_nparray(np_out)) mem_arg0 = ctypes.pointer(ctypes.pointer(rt.get_ranked_memref_descriptor(np_arg0))) mem_out = ctypes.pointer(ctypes.pointer(rt.get_ranked_memref_descriptor(np_out))) self._engine.invoke('main', mem_out, mem_arg0) return rt.ranked_memref_to_numpy(mem_out[0])
def build_compile_and_run_SpMM(attr: st.EncodingAttr, support_lib: str, compiler): # Build. module = build_SpMM(attr) func = str(module.operation.regions[0].blocks[0].operations[0].operation) module = ir.Module.parse(func + boilerplate(attr)) # Compile. compiler(module) engine = execution_engine.ExecutionEngine(module, opt_level=0, shared_libs=[support_lib]) # Set up numpy input and buffer for output. a = np.array( [[1.1, 0.0, 0.0, 1.4], [0.0, 0.0, 0.0, 0.0], [0.0, 0.0, 3.3, 0.0]], np.float64) b = np.array([[1.0, 2.0], [4.0, 3.0], [5.0, 6.0], [8.0, 7.0]], np.float64) c = np.zeros((3, 2), np.float64) mem_a = ctypes.pointer(ctypes.pointer(rt.get_ranked_memref_descriptor(a))) mem_b = ctypes.pointer(ctypes.pointer(rt.get_ranked_memref_descriptor(b))) mem_c = ctypes.pointer(ctypes.pointer(rt.get_ranked_memref_descriptor(c))) # Allocate a MemRefDescriptor to receive the output tensor. # The buffer itself is allocated inside the MLIR code generation. ref_out = rt.make_nd_memref_descriptor(2, ctypes.c_double)() mem_out = ctypes.pointer(ctypes.pointer(ref_out)) # Invoke the kernel and get numpy output. # Built-in bufferization uses in-out buffers. # TODO: replace with inplace comprehensive bufferization. engine.invoke('main', mem_out, mem_a, mem_b, mem_c) # Sanity check on computed result. expected = np.matmul(a, b) c = rt.ranked_memref_to_numpy(mem_out[0]) if np.allclose(c, expected): pass else: quit(f'FAILURE')
def create_sparse_tensor( filename: str, sparsity: Sequence[sparse_tensor.DimLevelType] ) -> Tuple[ctypes.c_void_p, np.ndarray]: """Creates an MLIR sparse tensor from the input file. Args: filename: A string for the name of the file that contains the tensor data in a COO-flavored format. sparsity: A sequence of DimLevelType values, one for each dimension of the tensor. Returns: A Tuple containing the following values: storage: A ctypes.c_void_p for the MLIR sparse tensor storage. shape: A 1D numpy array of integers, for the shape of the tensor. Raises: OSError: If there is any problem in loading the supporting C shared library. ValueError: If the shared library doesn't contain the needed routine. """ with ir.Context() as ctx, ir.Location.unknown(): module = _get_create_sparse_tensor_kernel(sparsity) module = ir.Module.parse(module) engine = compile_and_build_engine(module) # A sparse tensor descriptor to receive the kernel result. c_tensor_desc = _SparseTensorDescriptor() # Convert the filename to a byte stream. c_filename = ctypes.c_char_p(bytes(filename, "utf-8")) arg_pointers = [ ctypes.byref(ctypes.pointer(c_tensor_desc)), ctypes.byref(c_filename) ] # Invoke the execution engine to run the module and return the result. engine.invoke(_ENTRY_NAME, *arg_pointers) shape = runtime.ranked_memref_to_numpy(ctypes.pointer(c_tensor_desc.shape)) return c_tensor_desc.storage, shape