1 /*
2  * Licensed to the Apache Software Foundation (ASF) under one
3  * or more contributor license agreements.  See the NOTICE file
4  * distributed with this work for additional information
5  * regarding copyright ownership.  The ASF licenses this file
6  * to you under the Apache License, Version 2.0 (the
7  * "License"); you may not use this file except in compliance
8  * with the License.  You may obtain a copy of the License at
9  *
10  *   http://www.apache.org/licenses/LICENSE-2.0
11  *
12  * Unless required by applicable law or agreed to in writing,
13  * software distributed under the License is distributed on an
14  * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
15  * KIND, either express or implied.  See the License for the
16  * specific language governing permissions and limitations
17  * under the License.
18  */
19 
20 /*!
21  * \brief External function interface to rocBLAS libraries
22  * \file tags.h
23  */
24 #ifndef TOPI_CONTRIB_ROCBLAS_H_
25 #define TOPI_CONTRIB_ROCBLAS_H_
26 
27 #include "tvm/operation.h"
28 #include "topi/detail/extern.h"
29 
30 namespace topi {
31 namespace contrib {
32 using namespace tvm;
33 /*!
34 * \brief Create an op that multiplies lhs and rhs with rocBLAS
35 *
36 * \param lhs The left matrix operand
37 * \param rhs The right matrix operand
38 * \param transa Whether to transpose lhs
39 * \param transb Whether to transpose rhs
40 *
41 * \return The output tensor
42 */
rocblas_matmul(const Tensor & lhs,const Tensor & rhs,bool transa,bool transb)43 inline Tensor rocblas_matmul(const Tensor& lhs,
44                              const Tensor& rhs,
45                              bool transa,
46                              bool transb) {
47   auto n = transa ? lhs->shape[1] : lhs->shape[0];
48   auto m = transb ? rhs->shape[0] : rhs->shape[1];
49 
50   return make_extern(
51     { { n, m } }, { lhs->dtype }, { lhs, rhs },
52     [&](Array<Buffer> ins, Array<Buffer> outs) {
53       return call_packed({
54         Expr("tvm.contrib.rocblas.matmul"),
55         pack_buffer(ins[0]),
56         pack_buffer(ins[1]),
57         pack_buffer(outs[0]),
58         transa,
59         transb });
60     }, "C", "", {})[0];
61 }
62 
63 }  // namespace contrib
64 }  // namespace topi
65 
66 #endif  // TOPI_CONTRIB_ROCBLAS_H_
67