!10239 add StackPush and StackPop for aicpu

From: @yanzhenxiang2020
Reviewed-by: 
Signed-off-by:

mindspore/ccsrc/backend/kernel_compiler/aicpu/aicpu_util.h

@@ -44,6 +44,10 @@ constexpr auto kSeed2 = "seed2";
 constexpr auto kTopK = "TopK";
 constexpr auto kTopKV2 = "TopKV2";
 constexpr auto kStack = "Stack";
+constexpr auto kStackInit = "StackInit";
+constexpr auto kStackPush = "StackPush";
+constexpr auto kStackPop = "StackPop";
+constexpr auto kStackDestroy = "StackDestroy";
 constexpr auto kEditDistance = "EditDistance";
 constexpr auto kGatherD = "GatherD";
 constexpr auto kIdentity = "Identity";
@@ -55,8 +59,8 @@ constexpr auto kCustRunApi = "RunCpuKernel";
 constexpr auto kDropout2D = "Dropout2D";
 constexpr auto kDropout3D = "Dropout3D";
 const std::set<std::string> kCustAiCpuKernelOps{kIdentity};
-const std::set<std::string> kCacheKernelOps{kUpdateCache, kCacheSwapTable, kSubAndFilter,
-                                            kPadAndShift, kDropout3D,      kDropout2D};
+const std::set<std::string> kCacheKernelOps{kUpdateCache, kCacheSwapTable, kSubAndFilter, kPadAndShift, kDropout3D,
+                                            kDropout2D,   kStackInit,      kStackPush,    kStackPop,    kStackDestroy};
 
 struct AicpuParamHead {
   uint32_t length;         // Total length: include cunstom message

mindspore/ops/_op_impl/aicpu/__init__.py

@@ -70,4 +70,8 @@ from .fused_sparse_ftrl import _fused_sparse_ftrl_aicpu
 from .fused_sparse_proximal_adagrad import _fused_sparse_proximal_adagrad_aicpu
 from .meshgrid import _meshgrid_aicpu
 from .trans_data import _trans_data_aicpu
+from .stack_push_pop import _stack_init_aicpu
+from .stack_push_pop import _stack_push_aicpu
+from .stack_push_pop import _stack_pop_aicpu
+from .stack_push_pop import _stack_destroy_aicpu
 from .ctc_greedy_decoder import _ctc_greedy_decoder_aicpu

mindspore/ops/_op_impl/aicpu/stack_push_pop.py

@@ -0,0 +1,87 @@
+# Copyright 2020 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+
+"""StackPush and StackPop op"""
+from mindspore.ops.op_info_register import op_info_register, AiCPURegOp, DataType
+
+stack_init_op_info = AiCPURegOp("StackInit") \
+    .fusion_type("OPAQUE") \
+    .attr("index", "int") \
+    .get_op_info()
+
+stack_push_op_info = AiCPURegOp("StackPush") \
+    .fusion_type("OPAQUE") \
+    .input(0, "src", "required") \
+    .attr("index", "int") \
+    .dtype_format(DataType.U8_Default) \
+    .dtype_format(DataType.U16_Default) \
+    .dtype_format(DataType.U32_Default) \
+    .dtype_format(DataType.U64_Default) \
+    .dtype_format(DataType.I8_Default) \
+    .dtype_format(DataType.I16_Default) \
+    .dtype_format(DataType.I32_Default) \
+    .dtype_format(DataType.I64_Default) \
+    .dtype_format(DataType.F16_Default) \
+    .dtype_format(DataType.F32_Default) \
+    .dtype_format(DataType.F64_Default) \
+    .dtype_format(DataType.BOOL_Default) \
+    .get_op_info()
+
+stack_pop_op_info = AiCPURegOp("StackPop") \
+    .fusion_type("OPAQUE") \
+    .output(0, "dst", "required") \
+    .attr("index", "int") \
+    .dtype_format(DataType.U8_Default) \
+    .dtype_format(DataType.U16_Default) \
+    .dtype_format(DataType.U32_Default) \
+    .dtype_format(DataType.U64_Default) \
+    .dtype_format(DataType.I8_Default) \
+    .dtype_format(DataType.I16_Default) \
+    .dtype_format(DataType.I32_Default) \
+    .dtype_format(DataType.I64_Default) \
+    .dtype_format(DataType.F16_Default) \
+    .dtype_format(DataType.F32_Default) \
+    .dtype_format(DataType.F64_Default) \
+    .dtype_format(DataType.BOOL_Default) \
+    .get_op_info()
+
+stack_destroy_op_info = AiCPURegOp("StackDestroy") \
+    .fusion_type("OPAQUE") \
+    .attr("index", "int") \
+    .get_op_info()
+
+
+@op_info_register(stack_init_op_info)
+def _stack_init_aicpu():
+    """StackInit aicpu register"""
+    return
+
+
+@op_info_register(stack_push_op_info)
+def _stack_push_aicpu():
+    """StackPush aicpu register"""
+    return
+
+
+@op_info_register(stack_pop_op_info)
+def _stack_pop_aicpu():
+    """StackPop aicpu register"""
+    return
+
+
+@op_info_register(stack_destroy_op_info)
+def _stack_destroy_aicpu():
+    """StackDestroy aicpu register"""
+    return

mindspore/ops/operations/_inner_ops.py

@@ -15,6 +15,7 @@
 
 """Inner operators."""
 
+import numpy as np
 from ..._checkparam import Rel
 from ..._checkparam import Validator as validator
 from ... import context
@@ -882,3 +883,129 @@ class Centralization(PrimitiveWithInfer):
                'dtype': x_dtype,
                'value': None}
         return out
+
+
+class StackInit(PrimitiveWithInfer):
+    """
+    Create a stack that produces tensors in first-in last-out order.
+
+    After `StackInit`, a tensor can be pushed onto the stack using `StackPush`, and popped
+    at the top of the stack using `StackPop`. Finally, the stack should be destroyed with `StackDestroy`.
+
+    Args:
+        index (int): The index of the stack.
+
+    Supported Platforms:
+        ``Ascend``
+
+    Examples:
+        >>> x = Tensor(np.array([[1, 3], [2, 0]]))
+        >>> index = 0
+        >>> stack = ops.StackInit(index)
+        >>> push = ops.StackPush(index)
+        >>> pop = ops.StackPop(index, x.shape, x.dtype)
+        >>> destroy = ops.StackDestroy(index)
+        >>> stack()
+        >>> push(x)
+        >>> y = pop()
+        >>> destroy()
+        >>> print(y)
+        [[1 3]
+         [2 0]]
+    """
+    @prim_attr_register
+    def __init__(self, index=1):
+        """StackInit"""
+        validator.check_value_type("index", index, [int], self.name)
+
+
+class StackPush(PrimitiveWithInfer):
+    """
+    Push a tensor onto the stack.
+
+    Before `StackPush`, the stack should be created using `StackInit`.
+    Please refer to the usage in source code of `StackInit`.
+
+    Args:
+        index (int): The index of the stack.
+
+    Inputs:
+        - **input** (Tensor) - A tensor to be pushed onto the stack.
+
+    Supported Platforms:
+        ``Ascend``
+
+    Examples:
+        Please refer to the usage of `StackInit`.
+    """
+    @prim_attr_register
+    def __init__(self, index=1):
+        """StackPush"""
+        validator.check_value_type("index", index, [int], self.name)
+        self.init_prim_io_names(inputs=['input'], outputs=[])
+
+
+class StackPop(PrimitiveWithInfer):
+    """
+    Pop the tensor at the top of the stack.
+
+     Before `StackPop`, the stack should be created using `StackInit`.
+     Please refer to the usage in source code of `StackInit`.
+
+    Args:
+        index (int): The index of the stack.
+        shape (tuple): The shape of the tensor at the top of the stack.
+        dtype (mindspore.dtype): The type of the tensor at the top of the stack.
+
+    Outputs:
+        - **output** (Tensor) - The tensor at the top of the stack.
+
+    Supported Platforms:
+        ``Ascend``
+
+    Examples:
+        Please refer to the usage of `StackInit`.
+    """
+    @prim_attr_register
+    def __init__(self, index=1, shape=(1,), dtype=mstype.float32):
+        """StackPop"""
+        validator.check_value_type("index", index, [int], self.name)
+
+        validator.check_value_type('shape type', shape, [list, tuple], self.name)
+        validator.check_int(len(np.array(shape).shape), 1, Rel.EQ, "dim of shape", self.name)
+        for elem in shape:
+            validator.check_int(elem, 1, Rel.GE, 'shape element', self.name)
+            validator.check_value_type('type of shape element', elem, [int], self.name)
+
+        validator.check_type_name("dtype", dtype, (mstype.bool_,) + mstype.number_type, self.name)
+        self.shape = shape
+        self.dtype = dtype
+
+        self.init_prim_io_names(inputs=[], outputs=['output'])
+
+    def __infer__(self):
+        return {'shape': (list(self.shape)),
+                'dtype': (self.dtype),
+                'value': None}
+
+
+class StackDestroy(PrimitiveWithInfer):
+    """
+    Destroy the stack.
+
+     Before `StackDestroy`, the stack should be created using `StackInit`.
+     Please refer to the usage in source code of `StackInit`.
+
+    Args:
+        index (int): The index of the stack.
+
+    Supported Platforms:
+        ``Ascend``
+
+    Examples:
+        Please refer to the usage of `StackInit`.
+    """
+    @prim_attr_register
+    def __init__(self, index=1):
+        """StackDestroy"""
+        validator.check_value_type("index", index, [int], self.name)

tests/st/ops/ascend/test_aicpu_ops/test_stack.py

@@ -0,0 +1,128 @@
+# Copyright 2020 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+import numpy as np
+import mindspore.nn as nn
+import mindspore.context as context
+from mindspore import Tensor
+from mindspore.ops.operations import _inner_ops as P
+
+context.set_context(mode=context.PYNATIVE_MODE, device_target="Ascend")
+
+
+class Net(nn.Cell):
+    def __init__(self, index=0, shapes_and_types=None):
+        super(Net, self).__init__()
+        shapes_and_types.reverse()
+        self.init = P.StackInit(index)
+        self.push = P.StackPush(index)
+        self.pop = [P.StackPop(index, shape, dtype) for (shape, dtype) in shapes_and_types]
+        self.destroy = P.StackDestroy(index)
+
+    def construct(self, x1, x2, x3):
+        self.init()
+        self.push(x1)
+        self.push(x2)
+        self.push(x3)
+        y1 = self.pop[0]()
+        y2 = self.pop[1]()
+        y3 = self.pop[2]()
+        self.destroy()
+        return y1, y2, y3
+
+
+class NetTwoStack(nn.Cell):
+    def __init__(self, index=0, shapes_and_types=None):
+        super(NetTwoStack, self).__init__()
+        self.init_0 = P.StackInit(index)
+        self.push_0 = P.StackPush(index)
+        self.pop_0 = [P.StackPop(index, shape, dtype) for (shape, dtype) in shapes_and_types]
+        self.destroy_0 = P.StackDestroy(index)
+
+        index += 1
+        self.init_1 = P.StackInit(index)
+        self.push_1 = P.StackPush(index)
+        self.pop_1 = [P.StackPop(index, shape, dtype) for (shape, dtype) in shapes_and_types]
+        self.destroy_1 = P.StackDestroy(index)
+
+    def construct(self, x1, x2, x3):
+        self.init_0()
+        self.init_1()
+
+        self.push_0(x1)
+        self.push_1(x3)
+        y1 = self.pop_0[0]()
+        z1 = self.pop_1[2]()
+        self.push_0(x2)
+        self.push_0(x3)
+        self.push_1(x1)
+        self.push_1(x2)
+        y2 = self.pop_0[2]()
+        z2 = self.pop_1[1]()
+        y3 = self.pop_0[1]()
+        z3 = self.pop_1[0]()
+
+        self.destroy_0()
+        self.destroy_1()
+        return y1, y2, y3, z1, z2, z3
+
+
+def test_net():
+    x1 = Tensor(np.random.randn(4,).astype(np.float64))
+    x2 = Tensor(np.random.randn(4, 6).astype(np.float32))
+    x3 = Tensor(np.random.randint(100, size=(3, 4, 5)).astype(np.int32))
+
+    shapes_and_types = []
+    shapes_and_types.append((x1.shape, x1.dtype))
+    shapes_and_types.append((x2.shape, x2.dtype))
+    shapes_and_types.append((x3.shape, x3.dtype))
+
+    net = Net(2018, shapes_and_types)
+    y1, y2, y3 = net(x1, x2, x3)
+    print(x1)
+    print(x2)
+    print(x3)
+    print(y1)
+    print(y2)
+    print(y3)
+    assert np.array_equal(y1.asnumpy(), x3.asnumpy())
+    assert np.array_equal(y2.asnumpy(), x2.asnumpy())
+    assert np.array_equal(y3.asnumpy(), x1.asnumpy())
+
+
+def test_net_tow_stack():
+    x1 = Tensor(np.random.randn(4,).astype(np.float64))
+    x2 = Tensor(np.random.randn(4, 6).astype(np.float32))
+    x3 = Tensor(np.random.randint(100, size=(3, 4, 5)).astype(np.int32))
+
+    shapes_and_types = []
+    shapes_and_types.append((x1.shape, x1.dtype))
+    shapes_and_types.append((x2.shape, x2.dtype))
+    shapes_and_types.append((x3.shape, x3.dtype))
+
+    net = NetTwoStack(1998, shapes_and_types)
+    y1, y2, y3, z1, z2, z3 = net(x1, x2, x3)
+    print(x1)
+    print(x2)
+    print(x3)
+    print(y1)
+    print(y2)
+    print(y3)
+    assert np.array_equal(y1.asnumpy(), x1.asnumpy())
+    assert np.array_equal(y2.asnumpy(), x3.asnumpy())
+    assert np.array_equal(y3.asnumpy(), x2.asnumpy())
+
+    assert np.array_equal(z1.asnumpy(), x3.asnumpy())
+    assert np.array_equal(z2.asnumpy(), x2.asnumpy())
+    assert np.array_equal(z3.asnumpy(), x1.asnumpy())