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

/*******************************************************************************
* Copyright 2026 ZTE 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.
*******************************************************************************/

#include "common/c_types_map.hpp"
#include "common/memory_tracking.hpp"
#include "common/primitive_attr.hpp"
#include "common/type_helpers.hpp"
#include "common/utils.hpp"

#include "cpu/rv64/rvv_brgemm_conv_utils.hpp"

namespace dnnl {
namespace impl {
namespace cpu {
namespace rv64 {

using namespace dnnl::impl::status;
using namespace dnnl::impl::format_tag;
using namespace dnnl::impl::utils;
using namespace prop_kind;
using namespace data_type;

namespace brgemm_convolution_utils {

status_t init_conf(brgemm_conv_conf_t &jcp, const convolution_desc_t &cd,
        memory_desc_t &src_md, memory_desc_t &weights_md, memory_desc_t &dst_md,
        memory_desc_t &bias_md, primitive_attr_t &attr, int nthreads) {

    const cpu_isa_t isa = v;

    if (!one_of(cd.prop_kind, forward_training, forward_inference))
        return status::unimplemented;

    const memory_desc_wrapper src_d(&src_md);
    const memory_desc_wrapper weights_d(&weights_md);
    const memory_desc_wrapper dst_d(&dst_md);
    const memory_desc_wrapper bias_d(&bias_md);

    const bool with_groups = weights_d.ndims() == src_d.ndims() + 1;
    const int ndims = src_d.ndims();

    jcp.isa = isa;
    jcp.prop_kind = cd.prop_kind;
    jcp.ndims = ndims;
    jcp.nthr = nthreads;

    jcp.mb = src_d.dims()[0];
    jcp.ngroups = with_groups ? weights_d.dims()[0] : 1;
    jcp.ic = src_d.dims()[1] / jcp.ngroups;
    jcp.oc = dst_d.dims()[1] / jcp.ngroups;

    // Spatial dimensions.
    jcp.id = (ndims == 5) ? src_d.dims()[2] : 1;
    jcp.ih = (ndims >= 4) ? src_d.dims()[ndims - 2] : 1;
    jcp.iw = src_d.dims()[ndims - 1];
    jcp.od = (ndims == 5) ? dst_d.dims()[2] : 1;
    jcp.oh = (ndims >= 4) ? dst_d.dims()[ndims - 2] : 1;
    jcp.ow = dst_d.dims()[ndims - 1];

    // Kernel dimensions.
    const int goff = with_groups ? 1 : 0;
    jcp.kd = (ndims == 5) ? weights_d.dims()[goff + 2] : 1;
    jcp.kh = (ndims >= 4) ? weights_d.dims()[goff + ndims - 2] : 1;
    jcp.kw = weights_d.dims()[goff + ndims - 1];

    // Strides.
    jcp.stride_d = (ndims == 5) ? cd.strides[0] : 1;
    jcp.stride_h = (ndims >= 4) ? cd.strides[ndims - 4] : 1;
    jcp.stride_w = cd.strides[ndims - 3];

    // Dilations.
    jcp.dilate_d = (ndims == 5) ? cd.dilates[0] : 0;
    jcp.dilate_h = (ndims >= 4) ? cd.dilates[ndims - 4] : 0;
    jcp.dilate_w = cd.dilates[ndims - 3];

    // Padding.
    jcp.f_pad = (ndims == 5) ? cd.padding[0][0] : 0;
    jcp.t_pad = (ndims >= 4) ? cd.padding[0][ndims - 4] : 0;
    jcp.l_pad = cd.padding[0][ndims - 3];
    jcp.back_pad = (ndims == 5) ? cd.padding[1][0] : 0;
    jcp.b_pad = (ndims >= 4) ? cd.padding[1][ndims - 4] : 0;
    jcp.r_pad = cd.padding[1][ndims - 3];

    // Data types — only f32 supported.
    jcp.src_dt = src_d.data_type();
    jcp.wei_dt = weights_d.data_type();
    jcp.dst_dt = dst_d.data_type();
    jcp.bia_dt = bias_d.ndims() ? bias_d.data_type() : data_type::undef;

    if (!everyone_is(f32, jcp.src_dt, jcp.wei_dt, jcp.dst_dt))
        return status::unimplemented;
    if (jcp.bia_dt != data_type::undef && jcp.bia_dt != f32)
        return status::unimplemented;

    // Reject when output width or per-group input channels are too small
    // for per-kernel-position GEMM calls to be efficient.
    if (jcp.ow < 20 || jcp.ic < 16) return status::unimplemented;

    jcp.with_bias = jcp.bia_dt != data_type::undef;
    jcp.with_sum = attr.post_ops_.find(primitive_kind::sum) != -1;

    // Source and destination: NSPC (channel-last) only.
    const auto dat_tag = pick(ndims - 3, nwc, nhwc, ndhwc);

    if (src_d.format_kind() == format_kind::any)
        CHECK(memory_desc_init_by_tag(src_md, dat_tag));
    else if (!src_d.matches_one_of_tag(dat_tag))
        return status::unimplemented;

    if (dst_d.format_kind() == format_kind::any)
        CHECK(memory_desc_init_by_tag(dst_md, dat_tag));
    else if (!dst_d.matches_one_of_tag(dat_tag))
        return status::unimplemented;

    // Weights: plain spatial × IC × OC layout (same as gemm conv NSPC).
    const auto wei_tag = with_groups ? pick(ndims - 3, wigo, hwigo, dhwigo)
                                     : pick(ndims - 3, wio, hwio, dhwio);

    memory_desc_t want_wei_md = weights_md;
    CHECK(memory_desc_init_by_tag(want_wei_md, wei_tag));

    if (weights_md.format_kind == format_kind::any) {
        weights_md = want_wei_md;
    } else {
        if (want_wei_md != weights_md) return status::unimplemented;
    }

    if (bias_d.ndims() && bias_md.format_kind == format_kind::any)
        CHECK(memory_desc_init_by_tag(bias_md, format_tag::x));

    jcp.src_tag = dat_tag;
    jcp.dst_tag = dat_tag;
    jcp.wei_tag = wei_tag;

    return status::success;
}

} // namespace brgemm_convolution_utils

} // namespace rv64
} // namespace cpu
} // namespace impl
} // namespace dnnl