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 TVM_TOPI_CONTRIB_ROCBLAS_H_ |
25 | #define TVM_TOPI_CONTRIB_ROCBLAS_H_ |
26 | |
27 | #include <tvm/te/operation.h> |
28 | #include <tvm/topi/detail/extern.h> |
29 | |
30 | namespace tvm { |
31 | namespace topi { |
32 | namespace contrib { |
33 | |
34 | using namespace tvm::te; |
35 | /*! |
36 | * \brief Create an op that multiplies lhs and rhs with rocBLAS |
37 | * |
38 | * \param lhs The left matrix operand |
39 | * \param rhs The right matrix operand |
40 | * \param transa Whether to transpose lhs |
41 | * \param transb Whether to transpose rhs |
42 | * |
43 | * \return The output tensor |
44 | */ |
45 | inline Tensor rocblas_matmul(const Tensor& lhs, const Tensor& rhs, bool transa, bool transb) { |
46 | auto n = transa ? lhs->shape[1] : lhs->shape[0]; |
47 | auto m = transb ? rhs->shape[0] : rhs->shape[1]; |
48 | |
49 | return make_extern( |
50 | {{n, m}}, {lhs->dtype}, {lhs, rhs}, |
51 | [&](Array<Buffer> ins, Array<Buffer> outs) { |
52 | return call_packed({StringImm("tvm.contrib.rocblas.matmul" ), pack_buffer(ins[0]), |
53 | pack_buffer(ins[1]), pack_buffer(outs[0]), transa, transb}); |
54 | }, |
55 | "C" , "" , {})[0]; |
56 | } |
57 | /*! |
58 | * \brief Create an op that batch multiplies lhs and rhs with rocBLAS |
59 | * |
60 | * \param lhs The left matrix operand e.g. (batch_size, M, K) |
61 | * \param rhs The right matrix operand e.g. (batch_size, K, N) |
62 | * \param transa Whether to transpose lhs |
63 | * \param transb Whether to transpose rhs |
64 | * |
65 | * \return The output tensor |
66 | */ |
67 | inline Tensor rocblas_batch_matmul(const Tensor& lhs, const Tensor& rhs, bool transa, bool transb) { |
68 | auto batch_size = lhs->shape[0]; |
69 | auto n = transa ? lhs->shape[2] : lhs->shape[1]; |
70 | auto m = transb ? rhs->shape[1] : rhs->shape[2]; |
71 | |
72 | return make_extern( |
73 | {{batch_size, n, m}}, {lhs->dtype}, {lhs, rhs}, |
74 | [&](Array<Buffer> ins, Array<Buffer> outs) { |
75 | return call_packed({StringImm("tvm.contrib.rocblas.batch_matmul" ), pack_buffer(ins[0]), |
76 | pack_buffer(ins[1]), pack_buffer(outs[0]), transa, transb}); |
77 | }, |
78 | "C" , "" , {})[0]; |
79 | } |
80 | |
81 | } // namespace contrib |
82 | } // namespace topi |
83 | } // namespace tvm |
84 | |
85 | #endif // TVM_TOPI_CONTRIB_ROCBLAS_H_ |
86 | |