[topi] add ARM v8.2 udot (uint8) support (apache#3978)

* [topi] add ARM v8.2 udot (uint8) support

* fix test case

* fix common conv2d schedule

* add back fp32_time in test

* fix lint

* fix doc, add support for int32_lanes=4, signed int

* fix lint

* add ic_bn % 4 checker in schedule

Author: yzhliu

topi/python/topi/arm_cpu/__init__.py

@@ -3,6 +3,7 @@
 from . import conv2d
 from . import depthwise_conv2d
 from . import conv2d_transpose
+from . import conv2d_int8
 from . import bitserial_conv2d
 from . import bitserial_dense
 from . import injective

topi/python/topi/arm_cpu/conv2d_int8.py

@@ -0,0 +1,112 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you 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.
+# pylint: disable=invalid-name,unused-variable,unused-argument,no-member
+"""Conv2D int8 schedule on ARM"""
+
+import tvm
+from tvm import autotvm
+from .. import generic, tag
+from ..util import get_const_tuple
+from ..nn.conv2d import conv2d_NCHWc_int8
+from ..generic import conv2d as conv2d_generic
+from .. import nn
+from ..nn.conv2d import _get_workload as _get_conv2d_workload
+from .tensor_intrin import dot_int8_int8_int32
+
+
+def _get_default_config(cfg, data, kernel, strides, padding, out_dtype):
+    """
+    Get default int8 schedule config for the workload
+    """
+    wkl = _get_conv2d_workload(data, kernel, strides, padding, out_dtype)
+    is_kernel_1x1 = wkl.hkernel == 1 and wkl.wkernel == 1
+    if is_kernel_1x1:
+        conv2d_generic.fallback_schedule_cpu_1x1_int8(
+            cfg, wkl, int32_lanes=2, num_int8_elements=4)
+    else:
+        conv2d_generic.fallback_schedule_cpu_common_int8(
+            cfg, wkl, int32_lanes=2, num_int8_elements=4)
+
+
+@autotvm.register_topi_compute(conv2d_NCHWc_int8, ['arm_cpu'], 'direct')
+def _declaration_conv_NCHWc_int8(cfg, data, kernel, strides,
+                                 padding, dilation, layout, out_layout, out_dtype):
+    # layout and out_layout are not used here,
+    # we keep them for debug convenience when dumping autotvm workload
+    n, ic_chunk, ih, iw, ic_bn = get_const_tuple(data.shape)
+    in_channel = ic_chunk * ic_bn
+
+    oc_chunk, ic_chunk, kh, kw, ic_bn, oc_bn, n_elems = get_const_tuple(kernel.shape)
+    num_filter = oc_chunk * oc_bn
+
+    # If no config was set, we can fallback to NCHW config.
+    if cfg.is_fallback:
+        _get_default_config(cfg, tvm.placeholder((n, in_channel, ih, iw), dtype=data.dtype),
+                            tvm.placeholder((num_filter, in_channel, kh, kw), dtype=kernel.dtype),
+                            strides, padding, out_dtype)
+    return nn.conv2d_NCHWc_int8_compute(data,
+                                        kernel,
+                                        strides,
+                                        padding,
+                                        dilation,
+                                        layout,
+                                        out_layout,
+                                        out_dtype)
+
+
+@autotvm.register_topi_schedule(generic.schedule_conv2d_NCHWc_int8, ['arm_cpu'], ['direct'])
+def _schedule_conv2d_NCHWc_int8(cfg, outs):
+    """Create schedule for tensors"""
+    s = tvm.create_schedule([x.op for x in outs])
+    scheduled_ops = []
+
+    def traverse(op):
+        """Traverse operators from computation graph"""
+        # inline all one-to-one-mapping operators except the last stage (output)
+        if tag.is_broadcast(op.tag):
+            if op not in s.outputs:
+                s[op].compute_inline()
+            for tensor in op.input_tensors:
+                if isinstance(tensor.op, tvm.tensor.ComputeOp) and tensor.op not in scheduled_ops:
+                    traverse(tensor.op)
+
+        if 'conv2d_NCHWc_int8' in op.tag:
+            conv_out = op.output(0)
+            kernel = conv_out.op.input_tensors[1]
+            data_vec = conv_out.op.input_tensors[0]
+            data = data_vec.op.input_tensors[0] \
+                if isinstance(data_vec.op, tvm.tensor.ComputeOp) and "pad" not in data_vec.op.tag \
+                else data_vec
+            if isinstance(data.op, tvm.tensor.ComputeOp) and "pad" in data.op.tag:
+                data_pad = data
+                data = data_pad.op.input_tensors[0]
+
+            args = [s, cfg, data_vec, conv_out, outs[0]]
+            # int8 conv kernel is 7-dim
+            _, _, kh, kw, _, _, _ = get_const_tuple(kernel.shape)
+            dtype = "uint" if data.dtype == "uint8" else "int"
+            if kh == 1 and kw == 1:
+                conv2d_generic.schedule_conv_NCHWc_cpu_1x1_int8(
+                    *args, int32_lanes=4, intrin=dot_int8_int8_int32(int32_lanes=4, dtype=dtype))
+            else:
+                conv2d_generic.schedule_conv_NCHWc_cpu_common_int8(
+                    *args, int32_lanes=4, intrin=dot_int8_int8_int32(int32_lanes=4, dtype=dtype))
+
+        scheduled_ops.append(op)
+
+    traverse(outs[0].op)
+    return s

topi/python/topi/arm_cpu/tensor_intrin.py

@@ -0,0 +1,110 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you 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.
+# pylint: disable=invalid-name,unused-variable,unused-argument,no-member
+"""Conv2D int8 schedule on ARM"""
+
+import tvm
+
+def dot_int8_int8_int32(int32_lanes, dtype='uint'):
+    """
+    Int8 dot product by every 4 elements using ARM v8.2 udot.
+    This function takes two arrays of int8 datatype -- data[4] and
+    kernel[int32_lanes][4] -- and computes a dot product of data[4] with every
+    4 elements of kernels, resulting in output[int32_lanes] of uint32 datatype.
+    The pseudo code is as follows.
+
+    .. code-block:: c
+
+        void dot_int8_int8_int32(int8 data[4], int8 kernel[16][4], int32 output[16]){
+            for (int i = 0; i < int32_lanes; i++){
+                out[i] = 0;
+                for (int k = 0; k < 4; k++){
+                    out[i] += data[k] * kernel[i][k]
+                }
+            }
+        }
+
+    Physically, the kernel array sits in a vector register and
+    the data[4] is broadcasted to another vector register. This
+    function returns a TensorIntrin that can be used to tensorize
+    a schedule.
+
+    Parameters
+    ----------
+    int32_lanes: int
+        How many int32/uint32 to produce
+    dtype: str, optional, {"uint", "int"}
+        Whether it works on unsigned int or signed int
+
+    Returns
+    -------
+    intrin : TensorIntrin
+        The ARM uint8 TensorIntrin that can be used in tensorizing schedule
+    """
+    num_int8_elements = 4  # 4 int8 elements in int32
+
+    data = tvm.placeholder((num_int8_elements,), dtype='%s8' % dtype, name='data')
+    kernel = tvm.placeholder((int32_lanes, num_int8_elements), dtype='%s8' % dtype, name='kernel')
+
+    k = tvm.reduce_axis((0, num_int8_elements), name='k')
+    C = tvm.compute((int32_lanes,),
+                    lambda i: tvm.sum(data[k].astype('%s32' % dtype) *
+                                      kernel[i, k].astype('%s32' % dtype),
+                                      axis=k), name="C")
+
+    a_buffer = tvm.decl_buffer(data.shape, dtype='%s8' % dtype, name="a_buffer",
+                               offset_factor=1,
+                               strides=[1])
+    b_buffer = tvm.decl_buffer(kernel.shape, dtype='%s8' % dtype, name="b_buffer",
+                               offset_factor=1,
+                               strides=[tvm.var('s'), 1])
+
+    def _intrin_func(ins, outs):
+        def _instr(index):
+            ib = tvm.ir_builder.create()
+            if index == 1:
+                ib.emit(outs[0].vstore(0, tvm.const(0, '%s32x%d' % (dtype, int32_lanes))))
+                return ib.get()
+
+            dtype_a = '%s8x%d' % (dtype, num_int8_elements)
+            dtype_b = '%s8x%d' % (dtype, int32_lanes * num_int8_elements)
+            dtype_c = '%s32x%d' % (dtype, int32_lanes)
+
+            a_int8 = ins[0].vload([0], dtype_a)
+            re_int32 = tvm.call_pure_intrin('%s32' % dtype, 'reinterpret', a_int8)
+            # broadcast a
+            vec_ai32 = re_int32.astype(dtype_c)
+
+            vec_a = tvm.call_pure_intrin(dtype_b, 'reinterpret', vec_ai32)
+            vec_b = ins[1].vload([0, 0], dtype_b)
+            vec_c = outs[0].vload([0], dtype_c)
+
+            inst = 'udot' if dtype == 'uint' else 'sdot'
+            inst = 'llvm.aarch64.neon.%s.v%di32.v%di8' % (
+                inst, int32_lanes, int32_lanes * num_int8_elements)
+            vdot = tvm.call_llvm_intrin(dtype_c,
+                                        inst,
+                                        tvm.const(2, 'uint32'),
+                                        vec_c, vec_a, vec_b)
+            ib.emit(outs[0].vstore(0, vdot))
+            return ib.get()
+
+        # body, reset, update
+        return _instr(0), _instr(1), _instr(2)
+
+    with tvm.build_config(offset_factor=1, partition_const_loop=True):
+        return tvm.decl_tensor_intrin(C.op, _intrin_func, binds={data:a_buffer, kernel:b_buffer})