onednn-src 0.1.13

Source of oneAPI Deep Neural Network Library (oneDNN)
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203

/*******************************************************************************
* Copyright 2017 Intel Corporation
*
* 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.
*******************************************************************************/

#ifndef CPU_GEMM_X8S8S32X_CONVOLUTION_HPP
#define CPU_GEMM_X8S8S32X_CONVOLUTION_HPP

#include <memory>

#include "common/c_types_map.hpp"
#include "common/memory_tracking.hpp"
#include "common/primitive.hpp"

#include "cpu/platform.hpp"

#include "cpu/cpu_convolution_pd.hpp"

#include "cpu/gemm_convolution_utils.hpp"
#include "cpu/gemm_x8s8s32x_convolution_utils.hpp"

#include "cpu/gemm/gemm.hpp"
#include "cpu/zero_point_utils.hpp"

namespace dnnl {
namespace impl {
namespace cpu {

struct gemm_x8s8s32x_convolution_fwd_t : public primitive_t {
    struct pd_t : public cpu_convolution_fwd_pd_t {
        using cpu_convolution_fwd_pd_t::cpu_convolution_fwd_pd_t;

        DECLARE_COMMON_PD_T(src_md()->data_type == data_type::u8
                        ? IGEMM_S8U8S32_IMPL_STR
                        : IGEMM_S8S8S32_IMPL_STR,
                gemm_x8s8s32x_convolution_fwd_t, USE_GLOBAL_SCRATCHPAD);

        status_t init(engine_t *engine) {
            using namespace data_type;
            using skip_mask_t = primitive_attr_t::skip_mask_t;
            const auto dst_type = dst_md(0)->data_type;

            VDISPATCH_CONV(
                    DNNL_CPU_THREADING_RUNTIME != DNNL_RUNTIME_THREADPOOL,
                    VERBOSE_UNSUPPORTED_THREADPOOL_RUNTIME);
            VDISPATCH_CONV(is_fwd(), VERBOSE_BAD_PROPKIND);
            VDISPATCH_CONV(set_default_alg_kind(alg_kind::convolution_direct),
                    VERBOSE_BAD_ALGORITHM);
            VDISPATCH_CONV(
                    utils::one_of(dst_md()->data_type, f32, bf16, s32, s8, u8),
                    VERBOSE_UNSUPPORTED_DT_CFG);

            VDISPATCH_CONV(utils::one_of(src_md()->data_type, s8, u8),
                    VERBOSE_UNSUPPORTED_DT);
            VDISPATCH_CONV(
                    weights_md()->data_type == s8, VERBOSE_UNSUPPORTED_DT);
            VDISPATCH_CONV(IMPLICATION(with_bias(),
                                   utils::one_of(weights_md(1)->data_type, f32,
                                           bf16, s32, s8, u8)),
                    VERBOSE_UNSUPPORTED_DT_CFG);

            VDISPATCH_CONV(!has_zero_dim_memory(), VERBOSE_EMPTY_TENSOR, "");

            VDISPATCH_CONV(attr()->has_default_values(skip_mask_t::scales
                                           | skip_mask_t::zero_points
                                           | skip_mask_t::post_ops
                                           | skip_mask_t::sum_dt,
                                   dst_type),
                    VERBOSE_UNSUPPORTED_ATTR);

            VDISPATCH_CONV(attr()->post_ops_.check_sum_consistency(dst_type,
                                   /* is_int8 */ true),
                    VERBOSE_UNSUPPORTED_POSTOP);
            CHECK(attr_scales_ok());
            CHECK(attr_zero_points_ok());

            auto scratchpad = scratchpad_registry().registrar();

            // TODO: make `init_conf` assign initialized object to `jcp_`
            jcp_ = conv_gemm_conf_t();
            CHECK(jit_gemm_convolution_utils::init_conf(jcp_, scratchpad,
                    *desc(), src_md_, weights_md_, dst_md_, bias_md_, attr_,
                    dnnl_get_max_threads()));

            VDISPATCH_CONV(gemm_x8s8s32x_convolution_utils::post_ops_ok(
                                   attr()->post_ops_, &dst_md_),
                    VERBOSE_UNSUPPORTED_POSTOP);

            return status::success;
        }

        conv_gemm_conf_t jcp_ = utils::zero<decltype(jcp_)>();
    };

    gemm_x8s8s32x_convolution_fwd_t(const pd_t *apd) : primitive_t(apd) {}

    status_t init(engine_t *engine) override {
        CHECK(safe_ptr_assign(pp_ker_, pp_ker_t::create(pd(), pd()->jcp_)));
        return (pp_ker_) ? pp_ker_->create_kernel() : status::success;
    }

    status_t execute(const exec_ctx_t &ctx) const override {
        return execute_forward(ctx);
    }

private:
    const pd_t *pd() const { return (const pd_t *)primitive_t::pd().get(); }
    status_t execute_forward(const exec_ctx_t &ctx) const;
    status_t execute_forward_thr(const int ithr, const int nthr,
            const char *src_base, const int8_t *wei_base, const char *bia_base,
            void *dst_base, const float *scales, const float *dst_scales,
            const zero_point_call_params_t &zp,
            const memory_tracking::grantor_t &scratchpad,
            const void *post_ops_binary_rhs_arg_vec,
            const exec_ctx_t &ctx) const;

    using pp_ker_t = gemm_x8s8s32x_convolution_utils::pp_ker_t;
    std::unique_ptr<pp_ker_t> pp_ker_;
};

struct gemm_x8s8s32x_convolution_bwd_data_t : public primitive_t {
    struct pd_t : public cpu_convolution_bwd_data_pd_t {
        using cpu_convolution_bwd_data_pd_t::cpu_convolution_bwd_data_pd_t;

        DECLARE_COMMON_PD_T(diff_dst_md()->data_type == data_type::u8
                        ? IGEMM_S8U8S32_IMPL_STR
                        : IGEMM_S8S8S32_IMPL_STR,
                gemm_x8s8s32x_convolution_bwd_data_t, USE_GLOBAL_SCRATCHPAD);

        status_t init(engine_t *engine) {
            using namespace data_type;

            VDISPATCH_CONV(
                    DNNL_CPU_THREADING_RUNTIME != DNNL_RUNTIME_THREADPOOL,
                    VERBOSE_UNSUPPORTED_THREADPOOL_RUNTIME);
            VDISPATCH_CONV(desc()->prop_kind == prop_kind::backward_data,
                    VERBOSE_BAD_PROPKIND);
            VDISPATCH_CONV(utils::one_of(diff_dst_md()->data_type, s8, u8),
                    VERBOSE_UNSUPPORTED_DT);
            VDISPATCH_CONV(
                    weights_md()->data_type == s8, VERBOSE_UNSUPPORTED_DT);
            VDISPATCH_CONV(utils::one_of(diff_src_md()->data_type, f32, bf16,
                                   s32, s8, u8),
                    VERBOSE_UNSUPPORTED_DT);
            VDISPATCH_CONV(IMPLICATION(with_bias(),
                                   utils::one_of(weights_md(1)->data_type, f32,
                                           bf16, s32, s8, u8)),
                    VERBOSE_UNSUPPORTED_DT_CFG);
            VDISPATCH_CONV(set_default_alg_kind(alg_kind::convolution_direct),
                    VERBOSE_BAD_ALGORITHM);

            VDISPATCH_CONV(!has_zero_dim_memory(), VERBOSE_EMPTY_TENSOR, "");
            VDISPATCH_CONV(attr()->has_default_values(
                                   primitive_attr_t::skip_mask_t::scales),
                    VERBOSE_UNSUPPORTED_ATTR);
            CHECK(attr_scales_ok());

            auto scratchpad = scratchpad_registry().registrar();

            // TODO: make `init_conf` assign initialized object to `jcp_`
            jcp_ = conv_gemm_conf_t();
            return jit_gemm_convolution_utils::init_conf(jcp_, scratchpad,
                    *desc(), diff_src_md_, weights_md_, diff_dst_md_, bias_md_,
                    attr_, dnnl_get_max_threads());
        }

        bool support_bias() const override { return true; }

        conv_gemm_conf_t jcp_ = utils::zero<decltype(jcp_)>();
    };

    gemm_x8s8s32x_convolution_bwd_data_t(const pd_t *apd) : primitive_t(apd) {}

    status_t execute(const exec_ctx_t &ctx) const override {
        return execute_backward_data(ctx);
    }

private:
    status_t execute_backward_data(const exec_ctx_t &ctx) const;
    status_t execute_backward_data_thr(const int ithr, const int nthr,
            const char *diff_dst_base, const int8_t *wei_base,
            const char *bia_base, char *diff_src_base,
            const memory_tracking::grantor_t &scratchpad,
            const exec_ctx_t &ctx) const;
    const pd_t *pd() const { return (const pd_t *)primitive_t::pd().get(); }
};

} // namespace cpu
} // namespace impl
} // namespace dnnl

#endif